Population Momentum and Population Aging in Asia and Near-East Countries
August, 2000
STUDY TITLE: Population Momentum and Population Aging in Asia and Near-East Countries
SUBCONTRACTOR NAME: East-West Center PROJECT NAME: The POLICY Project PROJECT NUMBER; 936-3078 PRIME CONTRACT: CCP-C-00-95-00023-04 SUBCONTRACT NO: 5402.50E.EWC.01
INVESTIGATORS: Andrew Mason Sang-Hyop Lee Gerard Russo
This study is being supported by the Strategic and Economic Analysis Office, Asia/Near East Bureau, USAID. For additional information, contact Andrew Mason at [email protected]. Foreword
The objectives of the study are to document the rapid growth of older populations in the region, to address key issues and topics related to aging, and to assess the implications for programs and policies. The study takes a broad look at population aging in Asia and a more detailed look at seven Asia/Near East countries: the Philippines, Thailand, Indonesia, India, South Korea, Bangladesh, and Egypt. In this study, Asia includes East, Southeast, and South Asia, but not Western Asia. The included countries are shaded on the map that follows. This study was undertaken at the request of and with the support of the Strategic and Economic Analysis Office, Asia/Near East Bureau of USAID. Funding for this report was made available by the POLICY Project under USAID Contract No. CCP-C-00-95-00024-04. The POLICY Project is implemented by The Futures Group International in collaboration with Research Triangle Institute and The Centre for Development and Population Activities. The authors appreciate the support and useful suggestions provided by Koki Agarwal, Phil Estermann, James Gribble, Charles Llewellyn, Jerre Manarolla, Christopher McDermott, Robert Retherford, Zynia Rionda, John Stover, and Hania Zlotnik. We appreciate the able assistance provided by Piya Sereevinyayut and Ann Takeyasu.
Honolulu 2000
2 Table of Contents
I. Overview
II. Aging and Population Change
A. Population projections to 2050: The UN’s 1998 revision
B. Labor force projections to 2050
C. Demographic characteristics of the older population and labor force
D. Demographics and the support system for the elderly
1. Marital status
2. Population dependency ratio and the economic support ratio
E. Sources of uncertainty
1. Fertility trends
2. Mortality trends
3. Alternative variants
4. HIV/AIDS
5. Immigration
6. Asia’s economic crisis
III. The Implications of Aging and Policy Alternatives
A. Aging and economic growth in Asia
1. Aging and the economic support ratio
2. Aging and output per worker
3. Aging, saving, and wealth
4. Concluding observations on aging and economic growth
3 B. Health and health-care policy
1. National health expenditures
2. Aging and health expenditures
3. Income and health expenditures
4. Health care financing systems
a. Egypt
b. India
c. Japan
d. Philippines
e. South Korea
f. Thailand
5. Policy issues in health care financing design
a. Risk-sharing
b. Informational imperfections
c. Universal coverage
d. Provider payment mechanisms and policy
e. Equity
f. Group size
g. Mandated coverage
h. Catastrophic/disability insurance
i. Cost containment
j. Employment-based insurance programs
k. Competition and regulation
4 C. Labor and retirement policy
1. Retirement trends
2. Causes of retirement
a. Labor supply effects
b. Labor demand and retirement
c. Rigidities in labor market
3. Concluding remarks: policy options
D. Old-Age support systems
1. Family support systems
a. The current situation
b. The future of family support systems
2. The advantages and disadvantages of family support systems
a. Enforceability: Will systems persist?
b. Risk-pooling
c. Administrative efficiency
d. Economic efficiency and growth
e. Rates of return
f. Transition costs and unfunded liabilities
3. Public support systems
a. Current situation
b. Should governments require universal participation in
comprehensive programs?
c. Financing: funded or pay-as-you-go?
5 d. Risk-pooling and old-age support systems
e. Governance: centrally or privately managed?
f. Conclusion
IV. Conclusions
References
Appendices
Appendix A. Population projections – technical notes
Appendix B. Economically active population estimates and projections
Appendix C. Description of methodology for projections of proportion widowed
6 List of Figures
Figure 1. Asia's age transition. Figure 2. Population pyramids for Asia. Figure 3. Population of Asia, fifteen-year age groups, 1950-2050. Figure 4. Male labor force of Asia; ten-year age group 1950-2050. Figure 5. Female labor force of Asia; ten-year age group 1950-2050. Figure 6. Male labor force participation rates of Asia by age group. Figure 7. Female labor force participation rates of Asia by age group. Figure 8. Male labor force transition of Asia. Figure 9. Female labor force transition of Asia. Figure 10. Projections of proportions widowed. Figure 11. Global fertility transition, UN 1998 revision, medium variant. Figure 12. Analysis of TFR projection, comparison of medium variant to global historical pattern. Figure 13. Comparison of high and low fertility variants to historical experience. Figure 14. Global mortality transition, UN population projections. Figure 15. Life expectancy transition, 1950-1995. Figure 16. Comparison of UN mortality projections to historical data. Figure 17. Economic support ratio, Southeast Asia. Figure 18. Summary of economic support ratios, Asia and selected countries. Figure 19. Wealth and output: Taiwan, US, and France. Figure 20. Saving rate: Taiwan, US, and France. Figure 21. Investment rates, 1960 versus 1990, countries of the world. Figure 22. Gross national saving, Japan 1885-1997 and forecasts. Figure 23. Rates of total expenditure on health by GDP. Figure 24. Burden of death, India, Japan and South Korea. Figure 25. Retirement hazard rates, Asian men 1950-2010. Figure 26. Relationship between activity rates of elderly and the percentage of employment in agriculture sector, 1990.
7 List of Tables
Table 1. Summary measures of aging for Asia, major regions, six Asian countries and Egypt. Table 2. Fertility assumptions, medium UN projections. Table 3. Projected life expectancy, ANE countries. Table 4. Total labor force, labor force aged 65 and over, regions of Asia and eight ANE countries. Table 5. Demographic characteristics of the older population, Asia. Table 6. Demographic characteristics of the older population, Asia and ANE countries. Table 7. Summary of Dependency Ratios and Economic Support Ratios. Table 8. Projected life expectancy among the study countries, Asia and the Near East, 1995- 2050. Table 9. Population 65 and older. Table 10. Growth rate, population 65 and older, three mortality variants. Table 11. Old age dependency ratios, regions of Asia and seven ANE countries. Table 12. Total dependency ratio, regions of Asia and seven ANE countries. Table 13. HIV prevalence in selected Asia and Near East Countries. Table 14. Migrant populations, proportion of foreign born in population, and the growth rate of migrant population; 1965, 1975, 1985 and 1990. Table 15. Net migration, net migration rate, population growth rate, and contribution of net migration to population growth. Table 16. Immigrants admitted to US per 1000 person in the sending country, 1996-97. Table 17. Macroeconomic indicators of selected Asian countries, 1994-1998. Table 18. National health expenditures as percentage of GDP, 1997, WHO estimates. Table 19. National health expenditures as percentage of GDP, World Bank estimates. Table 20. National health expenditure as percentage of GDP 1960-1996, G7 countries, Australia and South Korea. Table 21. Estimated income elasticities form health care expenditure data. Table 22. Tuberculosis among populations 60 and older, 1990. Table 23. Out-of-pocket payments as percentage of total health expenditure: South Korea Table 24. Average retirement age among men, OECD countries 1950-1995 Table 25. Median age at retirement, Asian males aged 40 and over 1950-2010 Table 26. Main reasons for leaving the job, men aged 55-64, EU Table 27. Statutory retirement age. Table 28. Living arrangements for elderly, selected Asian countries. Table 29. Non-family living arrangements for the elderly 65 and older in Asia. Table 30. Percentage of elderly living with children by parents’ age, ASEAN survey. Table 31. Proportion of older adults living alone by age and sex, selected Asian countries. Table 32. Household size, children and adults per household, Asian countries. Table 33. Types of schemes providing cash benefits to the aged, disabled, and/or survivors, 1999. Table 34. Coverage of OASDI programs. Table 35. Social security expenditures and contributions, selected years. Table 36. Social security expenditures as a percentage of GDP, circa 1993.
8 List of Appendix Tables
Table A1. Summary of Alternative Population Projections, medium fertility scenario, Asia. Table A2. Summary of Alternative Population Projections, high fertility scenario, Asia. Table A3. Summary of Alternative Population Projections, low fertility scenario, Asia. Table A4. Summary of Aging, alternative mortality scenarios, Asia. Table A5. Total Dependency Ratio of Asia. Table A6. Summary of Alternative Population Projections, medium fertility scenario, East Asia. Table A7. Summary of Alternative Population Projections, high fertility scenario, East Asia. Table A8. Summary of Alternative Population Projections, low fertility scenario, East Asia. Table A9. Summary of Aging, alternative mortality scenarios, East Asia. Table A10. Total Dependency Ratio of East Asia. Table A11. Summary of Alternative Population Projections, medium fertility scenario, Southeast Asia. Table A12. Summary of Alternative Population Projections, high fertility scenario, Southeast Asia. Table A13. Summary of Alternative Population Projections, low fertility scenario, Southeast Asia. Table A14. Summary of Aging, alternative mortality scenarios, Southeast Asia. Table A15. Total Dependency Ratio of Southeast Asia. Table A16. Summary of Alternative Population Projections, medium fertility scenario, S South Asia. Table A17. Summary of Alternative Population Projections, high fertility scenario, South Asia. Table A18. Summary of Alternative Population Projections, low fertility scenario, South Asia. Table A19. Summary of Aging, alternative mortality scenarios, South Asia. Table A20. Total Dependency Ratio of South Asia. Table A21. Summary of Alternative Population Projections, medium fertility scenario, South Korea. Table A22. Summary of Alternative Population Projections, high fertility scenario, South Korea. Table A23. Summary of Alternative Population Projections, low fertility scenario, South Korea. Table A24. Summary of Aging, alternative mortality scenarios, South Korea. Table A25. Total Dependency Ratio of South Korea. Table A26. Summary of Alternative Population Projections, medium fertility scenario, The Philippines. Table A27. Summary of Alternative Population Projections, high fertility scenario, The Philippines. Table A28. Summary of Alternative Population Projections, low fertility scenario, The Philippines. Table A29. Summary of Aging, alternative mortality scenarios, The Philippines. Table A30. Total Dependency Ratio of The Philippines. Table A31. Summary of Alternative Population Projections, medium fertility scenario, Thailand. Table A32. Summary of Alternative Population Projections, high fertility scenario, Thailand. Table A33. Summary of Alternative Population Projections, low fertility scenario, Thailand. Table A34. Summary of Aging, alternative mortality scenarios, Thailand.
9 Table A35. Total Dependency Ratio of Thailand. Table A36. Summary of Alternative Population Projections, medium fertility scenario, Indonesia. Table A37. Summary of Alternative Population Projections, high fertility scenario, Indonesia. Table A38. Summary of Alternative Population Projections, low fertility scenario, Indonesia. Table A39. Summary of Aging, alternative mortality scenarios, Indonesia. Table A40. Total Dependency Ratio of Indonesia. Table A41. Summary of Alternative Population Projections, medium fertility scenario, India. Table A42. Summary of Alternative Population Projections, high fertility scenario, India. Table A43. Summary of Alternative Population Projections, low fertility scenario, India. Table A44. Summary of Aging, alternative mortality scenarios, India. Table A45. Total Dependency Ratio of India. Table A46. Summary of Alternative Population Projections, medium fertility scenario, Bangladesh. Table A47. Summary of Alternative Population Projections, high fertility scenario, Bangladesh. Table A48. Summary of Alternative Population Projections, low fertility scenario, Bangladesh. Table A49. Summary of Aging, alternative mortality scenarios, Bangladesh. Table A50. Total Dependency Ratio of Bangladesh. Table A51. Summary of Alternative Population Projections, medium fertility scenario, Egypt. Table A52. Summary of Alternative Population Projections, high fertility scenario, Egypt. Table A53. Summary of Alternative Population Projections, low fertility scenario, Egypt. Table A54. Summary of Aging, alternative mortality scenarios, Egypt. Table A55. Total Dependency Ratio of Egypt. Table B1. Summary of labor force and labor force participation rates, Asia. Table B2. Summary of labor force and labor force participation rates, East Asia. Table B3. Summary of labor force and labor force participation rates, Southeast Asia. Table B4. Summary of labor force and labor force participation rates, South Asia. Table B5. Summary of labor force and labor force participation rates, Japan. Table B6. Summary of labor force and labor force participation rates, South Korea. Table B7. Summary of labor force and labor force participation rates, The Philippines. Table B8. Summary of labor force and labor force participation rates, Thailand. Table B9. Summary of labor force and labor force participation rates, Indonesia. Table B10. Summary of labor force and labor force participation rates, India. Table B11. Summary of labor force and labor force participation rates, Bangladesh. Table B12. Summary of labor force and labor force participation rates, Egypt.
10 OVERVIEW
The Year 2000 is a demographic watershed in Asia. After a century of growth, the number of children reached a peak in 1999 and, with the turn of the millennium, has begun a slow, gradual decline. Both the enormous, sustained baby boom that characterized the twentieth century and the end of that baby boom have enormous implications for population growth and age structure for the foreseeable future. Fertility decline has been substantial in many Asian countries; in some, fertility has dropped below two children per woman. Despite declining fertility, the region will experience substantial population growth during the coming decades because of what is known as population momentum. Simply put, most population growth will occur because large cohorts of children and youth will, over time, replace much smaller cohorts of middle-aged adults and elderly. The same forces will also lead to population aging with important implications for economic growth, social support systems, and public policy related to retirement, health care, and pensions.
The future of Asian economies is uncertain. During the last few decades, standards of living have risen in many countries, in some spectacularly so. But even the high-performing economies of East and Southeast Asia have experienced turmoil in recent years. China,
Singapore, and Taiwan have, to this point, maintained solid growth. South Korea now appears to be recovering strongly. Others, such as Indonesia and Thailand, continue to encounter difficulties. In many Asian countries, standards of living remain low. Cambodia, Myanmar, Viet
Nam, Bangladesh, India, Nepal and Pakistan all had per capita incomes of less than $500 in
1998. Although some have improved their economic performance in recent years, achieving adequate standards of living remains a distant and elusive goal in these countries.
11 Recent evidence shows that demographic conditions in Asia are broadly favorable to economic growth (Bloom and Williamson, 1998; Mason, 1999). But as the demographic transition proceeds, rapid growth of older populations that neither work nor save may slow rates of economic progress. If aging gathers force before Asia’s low-income countries achieve adequate standards of living, the obstacles to development will be all the higher. Hence, the next few decades are a critical period for the region.
As aging proceeds, policies towards health and the economic security of the elderly are taking on increased importance. The decisions made regarding these sectors will have far- reaching implications. With increasing difficulty, countries will have to weigh the interests of the young against the old, reconsider the responsibilities of the family vis-à-vis the state, and balance concerns for equity against those for economic growth.
The material presented in Section II discusses demographic change in the region in some detail. We present population projections recently prepared by the United Nations (UN 1998), examine the underlying assumptions, and consider important sources of uncertainty. In Section
III, we discuss the impact of aging on the macroeconomy, the health sector, labor and retirement policy, and old-age support systems.
12 Aging and Population Change
Population aging began in Asia during the 1970s. Before then, rapid growth in the number of children was producing a younger population. Between 1950 and 1975, the percentage of the population under age 15 rose from 37 to 40 percent and the mean age dropped from 21.9 to 19.7 years of age. By the mid-1970s, the trend towards a younger population had reversed itself. Growth in the number of children slowed relative to the number of working-age and the number of elderly (Figure 1). Currently, if UN projections are accurate, the percentage who are children has dropped to 30 percent of Asia's population while the percentage in the working ages has risen to sixty-four percent. The percentage elderly has increased gradually during this period and will continue to rise for the foreseeable future. By 2050, 17% of those living in Asia are expected to be 65 or older; 19% under the age of fifteen; and, 64% between the ages of 15 and 64. The median age will have reached 39 (UN, 1998, medium projections).
Figure 1. Asia's Age Transition
More detailed snapshots of Asia's projected age- and sex-structure are provided by the age pyramids for 2000, 2025, and 2050 (Figure 2). The age pyramid for 2000 is similar to those found in other relatively young populations. There is a broad base consisting of large numbers of children and a narrow top consisting of the relatively few numbers of elderly. In the past, new cohorts of ever-increasing size entered the population, enlarging the base. In the future, however, the base of the age pyramid is expected to be stable with population growth concentrated at older ages.
13 Figure 2. Population Pyramids for Asia.
The relative stability of Asia's young population is already apparent among the youngest age groups. In 2000, the 0-4-year-old cohort and the 5-9-year-old cohort have essentially the same population size, and neither cohort is as large as the 10-14-year-old cohort. In the future, entering cohorts are projected to be similar or somewhat smaller than preceding ones. The major demographic phenomenon will be a "filling out" of the pyramid at older ages. As this occurs, all but the oldest cohorts are projected to stabilize at around 300-350 million people per five-year age group. Between 2000 and 2025, 99% of Asia's projected population growth is accounted for by the population 30 and older and only 1% by the population under the age of 30. In contrast, the population under the age of 30 accounted for 70% of the growth between 1950 and 1975 and nearly 40% of the growth between 1975 and 2000.
The changes in Asia's population are being driven by three inter-related demographic phenomena. The first is a sustained "baby boom" that produced the largest cohort of youth in the past and, possibly, in the foreseeable future. The baby boom led to accelerated growth in the number of children between 1950 and the late 1970s and more modest growth until 1999. The second demographic event is the emergence of relative stability in the number of children. After more than a century of growth, the number of children in Asia is expected to begin a period of very slow, sustained decline. The third demographic event is the region's mortality transformation. Life expectancy at birth increased from 41 in the early 1950s to 60 by the early
1980s and is projected to reach 68 years for 2000-05.
The impact of these events on the regions age structure are traced out in Figure 3 which charts Asia's population from 1950 to 2050 separately for 15-year-age groups: 0-14, 15-29, 30-
14 44, 45-59, 60-74, and 75+. This representation of age-structure is advantageous in that it facilitates following cohorts over time. The impact of the "baby boom" is evident in the accelerated growth of the child population, those aged 0-14, between 1950 and 1980 and more gradual growth during the last two decades of the 20th Century. The child population more than doubled in size, increasing from about 500 million in 1950 to about 1.1 billion in 2000. Asia's baby boom was different than the post-World War II baby boom that occurred in the US and many other Western countries. It was much longer lasting, and it occurred for different reasons.
The baby boom of the West resulted from an increase in rates of childbearing. Asia's baby boom was primarily the result of a decline in infant and child mortality.
The baby boom had an enormous impact on age structure that will continue well into the
21st century. The first group of baby boomers reached young adulthood in 1965 producing accelerated growth in the young adult population (15-29). In 1980 the first Asian baby-boomers had their thirtieth birthday and in 1995 reached forty-five years of age. As baby boomers enter these age groups during the next few decades, the most rapid growth in Asia's population will be among those in the prime working ages (30-59). Growth in the older population will accelerate begin in 2010 (or 2015) when the first baby boomers turn 60 (or 65).
Figure 3. Population of Asia, Fifteen year age groups, 1950-2050
Asia's future age structure will be influenced as much by the near stability of future cohorts of children as by the rapid growth of the baby boom generation. The year 2000 will produce a generation of children in Asia, the Y2K-generation, that is smaller than the preceding one. As the Y2K-generation ages, growth in the number of young adults and prime-age adults
15 will stabilize and possibly begin to decline. As this occurs, the older cohorts, baby boomers, will continue to enlarge the absolute and relative numbers belonging to older age groups.
The impact of continuing changes in mortality on age structure is less apparent in Figure
3 than changes in size of cohorts of children. However, declining mortality at older ages will have an important impact. As life expectancy rises in the future, the gains in survival rates will be increasingly concentrated at older ages. As a consequence, older age groups grow more rapidly during their high growth period than do younger age groups. Likewise, once growth ceases older cohorts decline somewhat more gradually than do young cohorts. Thus, declining mortality in the future will reinforce the shift to an older population.
In percentage terms, the older population will be the most rapidly growing segment of
Asia's population during the first half of the 21st century. The average annual rate of growth for the 60-74 year old population is projected at 2.9% and the 75+ population at 3.4%. In contrast, the 0-14 population is projected to decline at an annual rate of 0.2%, the 15-29 population is projected to be essentially the same in 2050 as in 200, the 30-44 and 45-59 populations are projected to increase at 0.5% and 1.5% annually.
The aggregate patterns for Asia are dominated by its two most populous countries, China and India. However, the general trends and the demographic forces that are influencing regional trends are also operating in other Asian countries and Egypt. The speed and timing of the transition to an older population will vary considerably among the countries within the region. In general, the countries of East Asia are furthest along in the aging process, followed by Southeast
Asia, and South Asia. Japan and, then, Singapore have the oldest populations in Asia. Among
Asia's major countries, Pakistan has the youngest population (measured by the median age in
16 2000). There 42% are under the age of 15 and only 5% are 65 and older. Aging measures for the seven countries examined in detail in this study are reported in Table 1.
Table 1. Summary measures of aging for Asia, major regions, six Asian countries and Egypt.
Population Projections to 2050: The UN’s 1998 Revision
This study makes detailed use of population projections prepared by the United Nations
Population Division and released in 1998 (UN 1998). Any population projection is based on a set of assumptions about long-range trends in demographic variables. The reality that emerges during the coming decades may differ considerably from projected values depending on a variety of social, political, economic, and demographic forces. Political instability, new rounds of economic crisis, the emergence of new infectious diseases, and the more rapid spread of
HIV/AIDS could lead to substantial, unanticipated deterioration in mortality conditions and depressed levels of fertility. More optimistically, medical breakthroughs could lead to a substantial extension of life and more rapid increases in life expectancy than anticipated. The future course of fertility in post-transition societies is very uncertain. How low fertility will decline, how long it remains at sub-replacement levels, and whether new baby booms will occur is primarily a matter of speculation.
Despite these uncertainties, population projections provide an important and useful framework for thinking about the future. At this point, we rely single projection to describe the broad demographic trends in Asia. Below, sources of uncertainty and alternative projections are considered in some detail.
17 The methodology is summarized here, but the interested reader can find a more detailed explanation in Zlotnik, 1999. The methodology used by the United Nations requires estimates of the population by sex and age category in a base year (1995) and subsequent trends in age- specific rates of fertility, mortality, and international immigration from 1995 to 2050.
An estimate of the base year population in each country is obtained by revising and updating the most recent population census using available data on fertility, mortality, and, in principle, immigration. Decennial population censuses are conducted in most Asian countries in years ending with 0 or 1; hence, the most recent direct data on population were collected in 1990 or 1991. In most countries, more recent estimates of fertility and mortality are available from surveys and/or civil registration systems. The most comprehensive data are available for fertility and child mortality. Many countries do not have recent data on adult mortality. Few countries have reliable information on international immigration. Of the 32 countries of East, Southeast, and South Asia with populations of 150,000 or more, 21 have estimates of fertility available after
1993; 18 have estimated of child mortality available after 1993; 5 have estimates of adult mortality after 1993. Thirteen additional Asian countries had estimates of adult mortality for the
1990-93 period (Zlotnik, 1999, Table 1).
The UN Population Division prepares three sets of projections used here. They differ in their assumptions about future trends in fertility. The medium variant distinguishes three groups of countries. The first consists of countries with total fertility rates above replacement level (2.1 births per women). In these countries, the total fertility rate is projected to decline smoothly until it reaches 2.1 births per women at which time it remains constant throughout the remainder of the projection. The second group consists of countries with a total fertility rate between 1.5 and
2.1 births per women. In these, the fertility rate is projected to converge to 1.9 births per women.
18 The third group consists of countries with very low fertility, a TFR below 1.5 births per woman.
In these countries, the TFR is projected to rise to a target level of 1.7 births per woman.
Of the seven countries examined in this study, five belonged to the high fertility group.
Of these, Indonesia is projected to reach replacement level first, in 2000-05; India and Egypt five years later; and the Philippines and Bangladesh in 2010-15. Two countries, South Korea and
Thailand, are low fertility countries where the TFR is projected to increase to 1.9 births per woman during the first part of the 21st century (Table 2).
Table 2. Fertility assumptions, medium UN projections.
In the low-fertility variant, the total fertility rate for high fertility countries is projected to decline to 1.6 births per woman; for low fertility countries to 0.4 births per women below the target fertility level used in the medium-fertility variant. Thus, the TFR in South Korea and
Thailand are projected to decline to 1.5 births per woman. In the high-fertility variant, the total fertility rate for high fertility countries is projected to decline to 2.6 births per woman. The TFR for the low fertility countries is projected to rise to 0.4 births per woman above the target level, a
TFR of 2.3 in South Korea and Thailand (Zlotnik, 1999).
UN population projections are based on a single set of mortality assumptions, rather than alternative variants as employed to characterize fertility. In most countries, life expectancy at birth is expected to increase steadily, based on the recent or medium-term experience, but more slowly as high levels of life expectancy are reached. Countries with a life expectancy under 65, for example, are projected to gains 2-2.5 years in life expectancy per quinquennium. When life expectancy reaches 70, the gain per quinquennia is only 1.0 years. At higher levels, the increase
19 per quinquennium drops to 0.3 to 0.5 years of life. The mortality pattern is assumed to converge to an ultimate life table with a life expectancy of 82.5 for males and 87.5 for females, although no country reaches the ultimate life expectancy by the end of the projection period (Zlotnik,
1999). Mortality assumptions for 3 Asian countries, Cambodia, India, and Thailand, explicitly include the impact of HIV/AIDS. Details of the procedures followed are available in Zlotnik,
1999.
The projected life expectancy, at ten-year intervals, for each of the study countries is reported in Table 3. In every country a steady increase in life expectancy is anticipated. The most rapid projected increase occurs in countries with low levels of life expectancy in 1990-95,
Bangladesh, Indonesia, Egypt, India, and the Philippines. Slower increases are anticipated in
Thailand and South Korea. The projections anticipate substantial convergence in mortality conditions. By 2040-45, the difference between the lowest and the highest country, Bangladesh and South Korea, is only 5.4 years. In 1990-95, the difference was estimated at 15.3 years.
Table 3. Projected life expectancy, ANE countries.
The final component of the UN population projections is international migration. The information about international migration is limited and unreliable, but in most instances international migration is a relatively small component of population growth. In some instances, however, political instability or natural disasters have generated large-scale movements between countries. These are largely unpredictable in nature and no attempt is made to capture them in the UN projections, although the projections do anticipate the return of refugees to the country of origin in some instances. The seven study countries are all classified as population "exporters",
20 although Thailand is a net exporter only for the period 2000-05. Thereafter, no net international migration is projected. Net outflows by South Koreans are projected to decline and stabilize at zero by 2020-25. The other study countries are projected to experience net outflows on a continuing basis. As a percentage of its population, the Philippines has the largest projected net outflows, averaging 0.1% of the population per year between 2000 and 2050. Indonesia and
Bangladesh both have net outflows equal to 0.06% of their population per year. The absolute number of out-migrants from India exceeds that from the other study countries, but is only
0.01% of the population per year. Obviously, net migration is a relatively small component of population growth even in the Philippines. The impact on age structure may be substantially larger than these net rates suggest to the extent that the age distribution of immigrants is heavily concentrated in particular age groups. Data on the age distribution of projected migrants are not available by the UN. Hence, the extent to which immigration is influencing population aging in
Asia cannot be directly assessed. This is an issue, however, to which we return below.
A brief overview of population projections was provided above. Detailed data are presented in Appendix A.
Labor Force Projections to 2050
The previous section examines the impact of population momentum on population aging in ANE countries. Population momentum also had deep implications for the size and age distribution of the labor force (economically active population). The broad labor force trends for
Asia are summarized by Figures 1 and 2 which show both the historical and projected labor forces in ten-year age groups for males and females. In many respects, the labor force trends are similar to the population trends described above. This is hardly surprising as a key determinant
21 of the size of the labor force in any age group is the size of the population in that age group. The values presented here, unless otherwise indicated, are based on the UN’s medium variant population projections.
In some countries, the labor force trends differ from the population trends because of important changes in labor force participation rates (activity rates). As discussed in more detail below, the activity rates of three demographic groups, in particular, are changing in many countries. First, many teenagers and young adults are extending their time in school and delaying their entry to the labor force. Second, activity rates among women have increased as rates of childbearing have declined. Third, activity rates among older males have dropped.
Estimates and projections of participation rates primarily rely on the ILO’s Economically
Active Population Estimates and Projection (EAPEP) for 1996. The EAPEP provides estimates and projections of labor force participation rates by sex and five-year age groups for the period
1950-2010 at ten-year intervals and for 1995. These projections were expanded to 2050 by making additional assumptions, the details of which are provided in Appendix B. The participation rates were combined with UN population from the 1998 Revision to yield estimates and projections of the labor force by age and sex.1
Figure 4. Male labor force of Asia; ten year age group 1950-2050
Figure 5. Female labor force of Asia; ten year age group 1950-2050
The labor force transition for Asia is dominated by its two largest countries, China and
India. However, most ANE countries have similar experiences with slight variation in the timing
1 The 1996 EAPEP uses the 1998 edition of the UN population projections.
22 and speed of the transition. The region and ANE countries have experienced substantial labor force growth that has come in waves, affecting the youngest first followed by successively older age groups. This phenomenon is very similar to what we have seen in the previous section, and directly reflects the impact of population momentum. The size of the child labor force, those aged 10-19, accelerated during the 1960s and 1970s, reached a peak in 1980, entered a period of decline that is projected to continue.2 Growth in the size of the labor force in this age group ceased earlier than its corresponding population because of the rapid decrease in activity rates of the age group since 1980. The young-adult labor force, those aged 20-29, increased most rapidly during 1970’s and 1980’s and is projected to reach a peak in 2010. The prime-age male and female labor force, those aged 30 to 49, has been increasing rapidly for the last two decades and is projected to peak in 2030. An older labor force, those aged 60 and over, is relatively small and has increased slowly during the last 50 years. Its era of rapid increase is not expected to begin until 2010 or later. In percentage terms, the elderly labor force is projected to increase most rapidly during the next 50 years. However, declining activity rates among the elderly could slow this growth considerably.
The labor forces of Asia, sub-regions, seven Asian countries and Egypt are provided in
Table 4. Detailed data and figures are provided in Appendix B. Note that the broad changes experienced in ANE countries are similar to those of the region as a whole. There are differences in timing and magnitudes. South Korea and Thailand are further along in their transitions and their labor forces are expected to grow little or not at all between 2000 and 2050. The elderly labor forces are expected to grow substantially in both countries but less so than in India,
Bangladesh, Egypt or elsewhere. Japan is much further along in the transition and, unlike most
2 The activity rates for those in this age group are assumed to be constant after 2010 but one can well imagine continued decline in those rates with further reductions in the size of the child labor force.
23 ANE countries, population decline is anticipated by the UN projections. Although Japan’s elderly labor force is expected to grow modestly, its total labor force is projected to decline by 20 million workers between 2000 and 2050.
Table 4. Total labor force, labor force aged 65 and over, regions of Asia and eight ANE countries
Changes in labor force participation rates have an important impact on the size and key characteristics of the labor force. The changes that have occurred in Asia, shown in Figures 6 and
7, are similar to those that individual ANE countries have also experienced. Three general changes have been pervasive. First, adult men aged 25-54 are reducing their participation slightly, while adult women are increasing theirs. Second, young men and women aged 15-24 are postponing their entry in the labor market. Third, elderly men are withdrawing from the labor force at a younger age. The effects of these changes are to concentrate the labor force at prime ages and to increase the number of women as compared to men.
The inverse U shaped profile of activity rates by age for men (Figure 6) is typical of most countries. The activity rates of men aged 24-54 are slightly decreasing in some countries because more workers are being discouraged in weak labor markets. Younger workers aged 15-24 show declining activity rates mostly due to the increase in the age limit for compulsory schooling and the growing trend among the young for higher education. In some instances, young people are discouraged and not seeking work because they think no work is available for their skills.
Figure 6. Male labor force participation rates of Asia by age group
24 For men aged 55 and over, activity rates have declined substantially for the last 50 years and are projected to continue to decrease, although more gradually, in the future. This is mostly because they have been discouraged and have ceased seeking work as employment opportunities for this age group have shrunk. Rapid technological progress has made their skills and competencies obsolete, reducing their productivity. Employers appear to be reluctant to employ or even to retain older workers on the grounds that aging reduces productivity, and seniority- based remuneration raises the cost of employing older workers. Many older persons have lost their job opportunities because employment has shifted out of agriculture. Jobs in agriculture have been the most important source of employment for the elderly, and this sector is not subject to the labor market rigidities that exist in an urban setting. Rapid changes in educational systems might also have given middle aged and younger workers a competitive advantage over their older counterparts. In some countries, mandatory early retirement provisions have affected activity rates among the elderly. Finally, for some developed countries such as Japan and Korea, higher standards of living, an increased preference for retirement, and the development of pension systems have contributed to lower activity rates among older men.
In most Asian countries, pension programs are sufficiently small that they have had a modest influence on retirement. However, as Asian countries begin to implement new programs for the elderly, participation rates could decline more rapidly than projected. This possibility and other policy aspects of labor force activity by older adults are considered in more detail below.
The activity rate profile of women (Figure 7) is moving closer to that of men, although, even in 2010, women are considerably less likely to be in the labor force than are men. The increase in female participation, concentrated in the 25-59 age group, can be traced to a variety of important developments. Changes in family responsibilities, particularly delayed marriage and
25 the decline in childrearing responsibilities, have played a particularly important role. The development of social infrastructure such as child-care, preschool and elderly-care facilities; changes in the work organization and the development of part-time work; the growth of new industries that require workers with dexterity and intelligence rather than brute strength have all contributed to the increased feminization of the labor force. Improved survey procedures also measure women’s activities more accurately by asking women what they do rather than asking them for their profession.
The trend toward higher female participation in Asia is not replicated in the experience of
Bangladesh or Thailand. Women in both of these countries have much higher activity rates than to other countries in the region. As their levels are already so high, a small decrease is foreseen for next 50 years.
For younger women in Asia, the trend in activity rates is similar to that for younger men.
As they remain in school longer their participation rates have declined. The gender gap in participation has largely disappeared at this age as young women are as likely to be in the labor force as are young men. For women aged 55 participation rates are generally low, have changed very little, and are projected to remain relatively constant.
Figure 7. Female labor force participation rates of Asia by age group
Activity rates vary widely among countries or sub-regions, reflecting varying labor market conditions, industrial structures, educational systems, gender bias, and pension systems.
Two differences are particularly important. First, participation by elderly men varies widely.
Whereas most of countries in Southeastern and South-central Asia recorded participation rates
26 ranging from 50 percent to 60 percent at the mid-1990’s, Japan and South Korea recorded rates of around 35 percent. Second, there are differences in the pattern of female participation. The M- shaped women’s activity profile found in relatively developed countries in the region such as
Japan and Korea was characteristic of the female life cycle in the 1950’s through the 1970’s. The bimodal shape was the result of women leaving the work force at around the age of 25 years to marry and raise children; a proportion returning later at around 35 years of age when their children had entered school or older children could take care of younger ones. This M-shape is not characteristic of other countries probably because of such factors as the dominance of agriculture sector, and/or different ways of child rearing. Even where the M-shape is found, the increase in female activity rates is causing its gradual disappearance.
The changing age structure of the Asian labor force reflects changes both in activity rates and changes in the underlying population. Declining activity rates among the young are reinforced by declining population growth rates among the young. Declining activity rates among the elderly, however, are more than offset by the rapid growth of elderly populations throughout Asia. The resulting age transition of the Asia population is shown in detail in Figures
8 and 9.
The proportion of Asian young male and female workers aged 10 to 19 has experienced a sharp decline for the last two decades. Although India and Bangladesh still will have a relatively high proportion of young workers (10-19) for the next decade, they too will begin to experience the same shift occurring elsewhere in the region. Prior to 1990, Asia as a whole did not experience any change in the proportion of old labor force aged 50 and over, although in Japan the proportion of labor force aged 50 and over has already increased very substantially. Korea and Thailand are also beginning to experience an increase in the proportion 50 and older. In the
27 coming decades, both for Asia as a whole and for most ANE countries, the proportion of labor force aged 50 to 59 and aged 60 and over is projected to increase markedly. By 2050, almost 30 percent of the male labor force is projected to be 50 or older; a somewhat smaller proportion of the female labor force will be concentrated at these ages.
Figure 8. Male labor force transition of Asia
Figure 9. Female labor force transition of Asia
In conclusion, if the projections presented here prove to be accurate, declining labor force participation rates and stable or declining population will lead to a decline in the number of young workers. Continued increases in female labor force participation combined with a larger population in the prime working ages will lead to a significantly larger female labor force. The number of elderly workers will also grow substantially even though the proportion of elderly men participating in labor force is expected to decline.
Demographic Characteristics of the Older Population and Labor Force
Currently Asia’s older population is concentrated at younger ages but over time the older members will disproportionately increase their numbers. Of the population 55 or older in 2000, roughly half are between the ages of 55 and 64 and roughly half are 65 or older. Of those 65 or older, roughly two-thirds are aged 65-74 and the remainder are 75 and above. Between now and
2025 the age-distribution of Asia’s older population is relatively stable, but between 2025 and
2050 the percentage of the older population aged 75 and above is expected to increase substantially. By 2050, among men, the percentage 75 and above is projected to reach 23 percent
28 of the older population; among women, the percentage 75 and above is projected to reach 30 percent (Table 5).
Table 5. Demographic characteristics of the older population, Asia.
In general, older women outnumber older men; they are also concentrated at older ages than older men. The reason is that, in most countries, survival rates for women are higher than for men. For the population 55 and older in 2000, Asia has about 10 women for every 9 men. At older ages, the sex ratio is even more unbalanced. Among those 75 and older, there are approximately 10 women for every 7 men. This is a persistent feature of Asia’s population and is expected to change only modestly over the 50-year projection period.
The older labor force is concentrated at younger ages reflecting the decline in labor force participation rates at older ages. In 2000, about 70 percent of workers in the 55 and older age group were aged 55-64 and about 30 percent were 65 or older. A modest shift in the distribution toward older ages after 2025 is anticipated. Older men are much more likely to be in the labor force than older women. In 2000, there are about 150 older men for every 100 older women in the labor force. The dominance of men among older workers is even greater at later ages. Among those 65 and older, there were 250 men in the labor force for every 100 women. Again, little change is anticipated during the next fifty years.
The demographic characteristics of the older populations vary within Asia. In East Asia, a somewhat higher percentage of the older population is 75 or older. Over time the differences are expected to increase. By 2050, one-third of the older population is expected to be 75 or older in East Asia. In South Asia and Southeast Asia, about one-quarter will fall into the oldest age
29 group. Among ANE countries, Japan’s older population is much more heavily concentrated at the oldest age group than is true of other countries. By 2050, over 40 percent are projected to be
75 or older in Japan. Aging is less advanced in other ANE countries --- Egypt and Bangladesh, for example (Table 6).
Table 6. Demographic characteristics of the older population, Asia and ANE countries.
The gender balance also varies among the regions and countries of Asia. Currently,
Southeast Asia has a relatively low sex ratio, i.e., relatively few men, as compared with East and
South Asia. Among the individual ANE countries, the populations of Japan, South Korea,
Thailand, and Egypt are most heavily weighted towards women. Bangladesh, on the other hand, has an unusually high sex ratio. Over the projection period, these differences become attenuated, although Japan, South Korea, and Thailand continue to have a relatively low sex ratio among their older populations (Table 6).
The activity rates of the older population are projected to decline during the coming decades. In part, this reflects an earlier age at retirement and, in part, changes in the age distribution of the older population. The most rapid changes tend to be in the countries experiencing the most rapid aging: Japan, South Korea, and Thailand. As this occurs, the older population will depend more on transfers and their financial assets and less on labor earnings.
30 Demographics and the Support System for the Elderly
The demographic trends described will have a sustained and irreversible impact on the nature of the support system on which the elderly rely. In this section, we describe these changes in more specific terms.
Marital status
Many elderly rely on their marital partner both for personal and financial support. Over time the proportion of the elderly who are married can shift substantially because of changes in the proportion celibate, i.e., who never marry, the proportion who divorce, mortality rates, and the proportions who remarry after divorce or the death of a spouse. In most Asian countries relatively low percentages never marry and rates of divorce are low. In the future, these two factors may become more important. The proportions of young adults who have ever-married have declined precipitously in many Asian countries. The decline may be entirely the result of a preference for a later age at marriage, but it may be that the proportion who never marry will rise. Likewise, rates of divorce have also increased in many Asian countries and, if rates of remarriage are low, the impact on the support system could be considerable at some point in the future.
In contrast with the uncertain changes in celibacy and divorce, changes in mortality rates are having and will continue to have a large impact on the proportion of the elderly who are married. Thus, the proportion widowed is the focus of this section.
A substantial portion of the elderly and particularly elderly women are widowed. In the year 2000, for example, we estimate once women reach their late sixties (65-69) about half of all
Korean and Indonesian women and about 40 percent of all Thai and Filipino women are widows.
31 Among older women the proportions are even greater. Depending on the nature of the support system, the loss of a spouse can have a devastating impact on the marital partner who remains.
The proportions of elderly men who are widowed are much lower. Of men in their late sixties, for example, we estimate that fewer than 15 percent were widowed in the year 2000. Detailed values are provided in Appendix C.
Women are much more likely to be widowed than men for several reasons. First, men are subject to higher mortality rates and typically die at a younger age than do women. Second, wives are typically younger than their husbands, adding further to the number of years by which they can expect to outlive their husbands. Third, older women are relatively unlikely to remarry in Asia should their husband die. A final issue is whether or not the mortality of husbands and wives is independent or not. If the loss of a spouse greatly increases mortality risks among women, for example, we would expect to see lower proportions who are widowed. Although the evidence is indirect, the statistical analysis presented in the appendix suggests that wives do not face substantially elevated mortality risks at the loss of their husband. On the other hand, husbands who experience the loss of a wife may be subject to greater mortality risks.
As mortality rates decline in the future, the proportion of men and women at any age who are widowed will decline in the absence of offsetting changes. Changes in any of the factors mentioned above, the age gap between husbands and wives, the proportions remarrying, and the correlation between mortality of husbands and wives, could also influence the proportions married. The results presented here, however, are based on the assumption that these other factors will remain constant and that only mortality changes will influence the proportions widowed. A detailed discussion of the methodology is presented in Appendix C.
32 Projections of the proportion widowed have been prepared for four Asian countries for which the requisite data were available. The trends are similar in all four countries, suggesting that it may be safe to generalize from their experience. As expected, the age-profiles of the proportions widowed shift steadily downward during the projection period. The changes are very substantial. A rough characterization is that, for women, the proportions widowed drops in half at each age group between 2000 and 2050. For men, the percentage changes are even greater and the proportions of males who are widowed reach low levels by 2050 (Figure 10).
Figure 10. Projections of proportions widowed.
Although these changes are large, they are not out of keeping with trends in Japan where the proportions widowed dropped very rapidly between 1970 and 1995. The steep decline reflects the very substantial increase in the proportion of men and women surviving to advanced ages.
The changes in widowhood have potentially far-reaching, but complex, implications for the support system for the elderly. In the future, older men and women can expect to have a surviving spouse to a much later age. That spouse may be an important resource providing personal care, companionship, and financial resources. On the other hand, that spouse may be a burden and a drain on the limited human and financial resources of his or her elderly partner.
Population dependency ratio and the economic support ratio
In all economic settings, the existence of large populations that are not currently productive requires that economic resources be transferred from the productive population to the
33 non-productive population. The relative sizes of those two groups determines in part the relative size of the transfers or the “burden” of the non-productive on the productive population.
The non-productive population consists of two age groups, the young and the elderly. As described above, the demographic transition is accompanied by important changes in the relative sizes of the productive and non-productive populations. In many Asian populations, rapid fertility decline has led to a rapid drop in the relative size of the child population and, consequently, a substantial rise in the productive population relative to the non-productive population. As the transition has proceeded, however, the relative size of the elderly dependent population has increased, producing a decline in the productive relative to the non-productive population. This demographic phenomenon undermines the capacity of the productive population to support the non-productive population.
These demographic changes are measured in two ways. The first makes use of the population dependency ratio. This is a standard measure. The child dependency ratio is conventionally constructed as the ratio of the population 0-14 to the population 15-64. The old- age dependency ratio is constructed as the ratio of the population 65 and older to the population
15-64. The total dependency ratio is the sum of these two or the ratio of the dependent population (0-14 plus 65 and older) to the productive population (15-64).
The second measure is the economic support ratio, which is the ratio of the productive population to the consuming population. Often the productive population is calculated using weights to adjust for age-variation in the labor force participation rates and productivity of the working-age population. The consuming population is constructed using weights to adjust for age-variation in material “needs”. We use a simple variant of the economic support ratio here.
34 For the productive population we use the labor force unadjusted for age-variation in productivity; for the consuming population we count children aged 0-15 as 0.5 and all adults as 1.0.
Dependency ratios and the economic support ratio are provided for the regions of Asia and ANE countries in Table 7. The table summarizes the important trends highlighted above.
The major trends anticipated in the region are a decline in the child dependency ratio and a rise in the old age dependency ratio during the next fifty years.
In East Asia, the old age dependency ratio dominates the trend in the total dependency ratio that increases modestly between 2000 and 2025 and more substantially between 2025 and
2050. In Southeast and South Asia, the trend in the total dependency ratio is dominated by the child dependency ratio between 2000 and 2025. Thereafter, the total dependency ratio begins to rise in step with the old age dependency ratio.
The pattern for the economic support ratio is similar to the pattern for the total dependency ratio. They move inversely because consumers are in the numerator and producers in the denominator in the support ratio. In the year 2000, the number of workers per effective consumer varies from 0.52 in South Asia to 0.68 in East Asia, giving East Asia an enormous advantage. By 2050, the demographic advantage of East Asia disappears entirely with both East and South Asia reaching 0.55 workers per effective consumer. South Asia will be in a slightly better position to support its dependent population in 2050; East Asia in a much worse position.
Southeast Asia’s position is essentially unchanged though advantaged as compared with East or
South Asia.
Table 7. Summary of dependency ratios and economic support ratios.
35 Sources of Uncertainty
The UN has been projecting populations since 1950 and, hence, the reliability of its projections has been widely scrutinized. In the past, projections by the UN have differed substantially from actual population trends because estimates of the base year population have been in error and because trends in birth and death rates have differed substantially from the projected levels. In recent years, global birth rates and death rates have declined more rapidly than anticipated and, consequently, global population growth has been slower and population aging more rapid than projected (Keilman, 1999). Other population projection efforts also have had limited success. Even though the US offers a much richer source of data on population and vital events than is typical elsewhere, population projections prepared by the US Census Bureau and the US Social Security Administration have not proven to be very accurate (Lee, 1999).
Fertility trends
The methodology employed by the UN for projecting fertility is described in general terms above. The medium variant fertility transition projected by the UN is characterized in more concrete terms by Figure 11. This figure plots the decline in TFR during any five-year period against the TFR at the beginning of the period.3 The TFR in the typical high fertility country is projected to decline by about 0.4 to 0.5 births per woman every five years until the TFR drops to about 2.8. At that point the pace of decline slows rapidly and stops when fertility reaches replacement level, a TFR of 2.1. The typical country with a TFR below 1.7 experiences a rise in the TFR until it reaches 1.7. A country with a TFR between 1.7 and 1.9 experiences a rise in the
TFR until it reaches 1.9, and a country with a TFR between 1.9 and 2.1 experiences no change in
36 its TFR. There is some variation around the typical pattern, but the differences between countries that are experiencing rapid fertility transition (90th percentile) and a slow fertility transition
(10th percentile) are relatively small as compared with historical experience.
Figure 11. Global Fertility Transition, UN 1998 Revision, Medium Variant.
The medium variant fertility transition is compared with the most recent historical pattern for the world based on UN estimates of the TFR for 1985-90 and 1990-95. For high fertility countries, the median pace of fertility decline from the UN medium variant are quite similar to the historical values. The historical range (shown in Figure 12), however, is much greater than the projected range (shown in Figure 11). For example, the pace of fertility decline in countries with a TFR near 5 was 0.2 births per quinquennium for the 10th percentile and 1.2 births per quinquennium for the 90th percentile, a difference of 1 birth per quinquennia. In contrast, the range in the projections is less than 0.2 births.4
Figure 12. Analysis of TFR Projection, Comparison of Medium Variant to Global Historical
Pattern.
If the fertility transition of the future is similar to the fertility transition of the past, the
UN medium variant projections will accurately capture the typical experience of high fertility
3 Figure 4 is constructed using UN projections of the TFR from 1995-2000 to 2045-50 for all countries. The data were sorted by TFR and moving samples of 99 observations were employed to calculate the median change in the TFR and the 10th and 90th percentiles fat each different TFR level, calculated as the median TFR of each sample. 4 The historical pattern is constructed using methods similar to those followed for constructing Figure 4. The moving sample is much smaller (37 observations) because a much smaller number of observations are available.)
37 countries. However, in any five-year period, the experience of any country can be expected to deviate substantially from the typical pattern. Moreover, the medium variant projection may systematically understate the fertility decline for regions or the world, because the distribution of fertility decline in the historical data is highly skewed. In the past, a substantial number of countries experienced very rapid fertility decline a phenomenon that is not captured in the projected data.
For many purposes a time perspective longer than five years is appropriate. If the within country variation in fertility decline is not highly autocorrelated, the average fertility decline over a ten, fifteen, or twenty year period will be much closer to the median. This occurs because rapid fertility decline in one five-year period is balanced by slow fertility decline in another five-year period. One simple method of assessing this issue is to calculate the average decline per quinquennium over a ten-year period using data for 1980-85 and 1990-95. The 90th percentile is substantially lower, about 0.9 births per quinquennium for TFRs in the 5 to 6 range. This confirms that the average rate of change for longer periods tends to fall within narrower bans than the range for five years plotted in the figures presented here.
In the past, UN projections have systematically underestimated the pace of fertility decline because fertility decline has accelerated over time. If the trend to a more rapid fertility transition continues, then the projected fertility in 1998 Revision once again will be higher than the actual. In the absence of information about why fertility decline accelerated, however, it is difficult to anticipate whether or not such rapid fertility decline will occur and in which countries. Rapid fertility decline in the past few decades has been associated to a considerable extent with rapid social and economic development in East and Southeast Asia. Rapid fertility decline has also occurred in Bangladesh despite its low level of development. To a great extent
38 rapid fertility decline in other high fertility countries will depend to some extent, but not entirely, on whether they enjoy development success. In the past, a number of modeling efforts have attempted to tie demographic trends explicitly to changes in social and economic variables although this approach has largely fallen out of fashion. Recently, Sanderson has argued that incorporating information about socioeconomic variables can lead to improved population forecasts (Sanderson, 1999).
During the past few decades, accurately projecting the speed of fertility decline among high fertility countries has been particularly important because such a large percentage of the world's population lived in high fertility countries. As the demographic transition has continued, however, attention will increasingly focus on forecasting fertility in low fertility countries. Of the seven study countries examined here, for example, two (South Korea and Thailand) have below replacement fertility, four (Indonesia, Egypt, Bangladesh, and India) have TFRs projected at between 2 and 3 births per woman for 1995-2000, and the Philippines has a projected TFR of 3.2 for 1995-2000 (Table 2). How rapidly these countries reach low levels of fertility is becoming less of an issue, and whether or not fertility drops far below replacement fertility is becoming more important.
At low levels of fertility, the medium variant of the 1998 Revision differs substantially, in both quantitative and qualitative terms, from historical experience. The actual decline in fertility between 1985-90 and 1990-95 was substantially more rapid among low fertility countries than in the projections. Even at the lowest TFR captured by the historical median, a TFR of about 1.8, fertility is continuing to decline by about 0.1 births per five-year period. At low levels fertility decline slowed, but it did not stabilize in the typical country. Even if we examine the eleven countries with TFRs below 1.6 births per woman in 1985-90, five experienced a further decline
39 in fertility and six experienced a rise in fertility. Nothing in the historical experience of low fertility countries supports the assumption that high fertility countries will converge to replacement fertility nor that low fertility countries will experience a rise in fertility, although neither of these possibilities can be ruled out.
Standard practice, and the UN approach to uncertainty, is to offer alternative variants. As explained above, the UN provides two variants employing alternative assumptions about fertility.
The high and low fertility variants are compared to the recent historical fertility transition in
Figure 13. Note that these variants differ primarily with respect to fertility trends at low and moderate TFRs. The high fertility variant assumes that, among high fertility countries, the TFR will stabilize at 2.5. Thus, projected fertility decline is much slower than was the case in the recent past (or earlier for that matter). The low fertility variant more closely tracks historical experience. Given this variant, projected fertility continue to decline even at low levels. The pace of the decline is substantially more rapid than has been true in recent history.
Figure 13. Comparison of high and low fertility variants to historical experience.
Emphasizing the typical pattern, the median, in historical data tends to de-emphasize the wide range of historical experience at low fertility levels. Some countries with low fertility levels experienced much more substantial declines in fertility than is typical; some experienced substantial increases in their fertility levels. Moreover, the historical patterns presented in the figure are based on limited experience with low fertility. One cannot rule out the possibility that fertility rates will drop to or remain at very low levels or that they will rise the levels well above replacement. Lee’s (1993) analysis of US fertility trends employing a stochastic forecasting
40 model, for example, estimated a 95% confidence interval for average US TFR from 1990 to 2065 from a high of 2.7 births per woman to a low of 1.3 births per woman.
The implications of alternative fertility outcomes for population aging are discussed below, but before turning to that we examine uncertainty with regard to future mortality.
Mortality trends
The UN projected mortality transition is represented in Figure 14 using the same methodology employed to represent the fertility transition above. The typical high mortality country is projected to experience an increase in life expectancy at birth of about 2.5 years per quinquennium. Once life expectancy reaches about 60, the rate of increase slows gradually.
Countries with life expectancies exceeding 80 typically experience an increase of 0.4 years per quinquennium in the UN projections. Note that UN does not use alternative mortality variants; hence, the mortality pattern portrayed in Figure 14 applies to all variants. At low levels of life expectancy the rate of increase varies considerably among the countries; the range is vary narrow at higher levels of life expectancy.
Figure 14. Global Mortality Transition, UN Population Projections.
The mortality transition based on historical data for 1950-95 is represented in Figure 15.
There are apparent similarities and apparent differences between the historical data and the UN projections. The median rate of increase for countries with life expectancies between 45 and 60 are very similar; the variances are also similar for life expectancies in the 45 to 55 year range. At higher life expectancies the projected decline for the typical country is much smoother and
41 continuous than has been true in the past. The historical data shows a peculiar dip for countries with a life expectancy of around 70. For countries with a life expectancy in the 72-75 range the rate of decline is about one year per quinquennium with no clear indication of a slow down. The range is much greater in the historical data at high levels of life expectancy than in the UN projections.
Figure 15. Life Expectancy Transition, 1950-1995.
In Figure 16, the UN mortality projections are compared both to the full set of historical data available and to the most recent historical period, 1985-95. The historical patterns differ markedly at low levels of life expectancy. The gains were much smaller between 1985-90 and
1990-95 reflecting primarily the impact of the AIDS epidemic in sub-Saharan Africa. Although the UN projections explicitly incorporate the impact of the AIDS epidemic, they anticipate a return to a mortality transition that is similar to the experience for the 1950-95 period.
The UN projections do not incorporate the dip in the rate of increase that occurs at life expectancy of 70 in both historical patterns. The historical pattern reflects relatively stagnant mortality conditions in Eastern European countries. Whether change in that region of the world eventually will result in more rapid improvements in mortality remains to be seen, but UN projections appear to be optimistic on this score.
Figure 16. Comparison of UN mortality projections to historical data.
42 The UN projections also anticipate continued decline in the rate of change at high levels of life expectancy. Given the paucity of data at life expectancies so high, there is little direct empirical evidence regarding this issue. The historical patterns represented here do not extend beyond 75 years of life expectancy because the last observation is based on the 99 countries with the highest life expectancy in the world in 1985-90. If we reduce the size of the sample so as to focus on a select group of countries with exceptionally high levels of life expectancy we find some indication of slower increase. The 14 countries with life expectancies exceeding 77 (a median of 77.4) experienced a median increase in life expectancy of 0.83 years per quinquennium. Although their experience suggests a further slowdown, the rate of increase was greater than in the UN projections. At a median life expectancy of 77.4 the median increase projected by the UN is 0.65 years per quinquennium.
How do the differences between the historical pattern and the UN projections bear on the course of life expectancy in the countries that are the focus of this study? In 1990-95, life expectancy varied from a low of 56 in Bangladesh to a high of 71 in South Korea (Table 3).
Projected life expectancy in each of these countries is higher, but less so for South Korea and
Thailand, because the UN model projects more rapid increase in life expectancy near 70 than has been the case of the historical data. South Korea and, to a lesser extent, Thailand are most influenced by uncertainties about trends in life expectancy at high levels. If the slow down envisioned in the UN projections does not emerge, both of these countries would have higher life expectancies.
This issue is examined more formally in Table 8 which compares the projected increase in life expectancy with the increase that would occur if each country followed the median path in the historical data. The historical model is based on centered moving average of 99 observations
43 for countries with a population of 2 million or more. Data for 1950-95 are included. At the upper ranges of life expectancy the size of the moving sample was reduced, as necessary, to obtain estimates. The smallest sample included is 30 observations for which the median increase was
0.9 years per quinquennium. The median was similar for the highest 15 observations, but for the highest ten observations the median increase was approximately 0.5 years per quinquennium.
The UN methodology anticipates more rapid increases in life expectancy in Bangladesh,
Indonesia, Egypt, and the Philippines than does the historical model. The UN anticipates a slower increase in life expectancy in South Korea, Thailand, and India than does the historical model. The slower increases in UN projections of life expectancy in Thailand and India are due, in part, to the projected impact of AIDS in those countries.
Table 8. Projected life expectancy among the study countries, Asia and the Near East, 1995-
2050.
The differences between the average rate of increase in life expectancy in the UN projections and a historically based model are relatively modest (about 10%) and it may well be that the differences in the UN projections are entirely appropriate. However, the differences between the historical data and the UN projections highlight the possibility that political changes, the emergence of infectious diseases, medical advance, and other unforeseen events may have an important bearing on the course of life expectancy in the future. Moreover, individual countries may deviate substantially from the typical pattern. The historical data indicates a relatively wide range in the pace of mortality decline.
The UN projections do not provide alternative mortality variants that would allow an assessment of the impact on population aging of unusually rapid or unusually slow mortality
44 decline. For Asia, sub-regions of Asia, and the seven study countries from Asia and the Near
East, however, we have estimated two variants that allow such an assessment. This was accomplished by preparing alternative projections. In the high mortality variant, life expectancy is assumed to increase half as rapidly as in the UN projections. Hence, the projected life expectancy achieved in 2020-25 in the UN projections is not achieved until 2040-45 in the high mortality projection. In the low mortality variant, life expectancy is assumed to increase twice as rapidly as in the UN projections. Hence, the projected life expectancy achieved in 2040-45 in the
UN projections is achieved in 2020-25 in the low mortality projections. Preparing alternative projections in this manner is feasible using the data that are publicly available from the UN. A chief limitation is that the low mortality projections cannot be extended beyond 2020-25. There is no obvious way that this limitation can be circumvented without modifying and fully implementing the UN projection methodology, an effort that is beyond the scope of this study.
Before we present and discuss alternative variants, we should consider the difficulties of interpreting these variants. The choice of assumptions used to implement variants is to a great extent arbitrary and we can provide little guidance about the probability that vital rates will fall within the span captured by alternative scenarios. Likewise, we cannot attach probabilities to the alternative outcomes. Our rough assessment is that the span in mortality decline between the high and low variant is wide. Mortality increase twice as rapid as projected by the UN is greater than the 90th percentile in the historical data. Given historical experience, the probability that a country would experience such rapid mortality decline consistently over a twenty-five year period must be substantially less than 0.1. Experiencing mortality decline that is half a rapid as in the UN projections is roughly comparable to the 10th percentile in the historical data, but again the probability of a country experiencing such a slow rate of mortality change for a sustained
45 period of time would be much lower. Balanced against these observations is the possibility of large unforeseen events that have a large and sustained impact on mortality and the uncertainties about the pace of mortality change at high levels of life expectancy.
Alternative variants
Detailed population projections for Asia and the seven study countries are provided in the appendix tables. Here we briefly describe results from the alternative scenarios.
The size of the population 65 and older is not influenced by fertility variation for 65 years. Thus, the number of elderly is identical in the three fertility variants produced by the
United Nations. The impact of mortality variation on the size of the elderly population is reported in Table 9. Irrespective of the mortality variant, the elderly population will increase substantially by 2025 and between 2025 and 2050. In all cases, the elderly population at least doubles between 2000 and 2025. In Southeast Asia and South Asia, the projected population 65 and older doubles again between 2025 and 2050. The size of the elderly population is affected by the mortality variant. If low mortality conditions prevail between 2000 and 2025, the population
65 and older is approximately 20% larger than if high mortality conditions hold.
Table 9. Population 65 and older.
The population growth rates of the population 65 and older reinforce the conclusion that rapid growth of the elderly population during the next fifty years can be safely assumed, even if it is not certain (Table 10). The most rapid growth in the elderly population is likely to occur in
Southeast Asia, but this conclusion is sensitive to the mortality variant. If mortality rates decline
46 slowly in Southeast Asia and rapidly in South Asia, then South Asia would experience the most rapid growth in the elderly population. Of the seven countries for which separate projections have been prepared, the Philippines is projected to experience the most rapid growth in the elderly population between 2000 and 2025 and Bangladesh is projected to experience the most rapid growth during the second quarter of the 21st century. Most countries are projected to experience somewhat slower growth in their elderly populations during the second quarter than the first quarter of the 21st century. The decline is particularly marked in South Korea. The mortality variant has a substantial impact on the rate of growth of the elderly population. During the first quarter century, the difference in the low mortality variant and the high mortality variant is consistently more than 0.5 percentage points but less than 0.8 percentage point.
Table 10. Growth rate, population 65 and older, three mortality variants.
The old age dependency ratio, the ratio of the population 65 and older to the population
15-64, is influenced both by mortality variation and, after 15 years, by fertility variation. The ratios are available for the nine combinations of fertility and mortality projected. Table 11 presents only the variants which produce the lowest old age dependency ratio (var1: high fertility, high mortality), the UN's medium fertility projection (var5), and the highest old age dependency ratio (var9: low fertility and low mortality until 2025; var6: low fertility and medium mortality after 2025). In 2025 the highest and lowest dependency ratios differ by roughly 20 percent. By 2050 the range widens considerably even though variant 9 is not available. The medium mortality, low fertility variant old age dependency ratio is 50 percent higher or substantially more for variant 6 than for variant 1.
47 Table 11. Old age dependency ratios, regions of Asia and seven ANE countries.
The demographic trends that produce the most extreme values of the old age dependency ratio do not necessarily produce the most extreme values of the total dependency ratio. Low fertility and low mortality rates produce the highest old age dependency ratio, but not the highest overall dependency ratio (Table 12). The reason is that low fertility yields a population with fewer in the working ages, but also fewer children. In the variants analyzed here, lower fertility generally produced a lower overall dependency ratio even though it produced a higher old age dependency ratio. Mortality variation of the type evaluated here generally has the same impact on total and old age dependency ratios because lower mortality rates produce more surviving children and more surviving elderly. Hence, low mortality is associated both with high old age dependency and high total dependency. In both 2025 and 2050, low fertility and high mortality consistently produced the lowest overall dependency ratio (variant 3) and high fertility and low mortality consistently produced the highest overall dependency ratio (variant 7 in 2025; variant 4 in 2050).
The changes in the overall dependency ratios between now and 2025 are very sensitive to the demographic assumption. In 2025, the highest total dependency ratio exceeded the lowest total dependency ratio by between 10 and 18 percentage points, depending on the country or region. The old age dependency ratios differed by only 2 to 3 percentage points across variants.
By 2050, however, the value of the old age dependency ratio differs considerably from one variant to the next.
48 Table 12. Total dependency ratio, regions of Asia and seven ANE countries.
HIV/AIDS
The emergence of HIV/AIDS has greatly increased uncertainty about population trends, especially in sub-Saharan Africa. To this point, many Asian countries have experienced much lower rates of exposure. However, Thailand, other countries in Indochina, southern China, and
India are experiencing serious epidemics. The emergence of HIV/AIDS in other Asian countries in the next few decades could have a serious demographic impact that will influence the course of aging and support systems.
UN Population Projections incorporate the impact of the HIV/AIDS epidemic in three
Asian countries of which two, Thailand and India, are subject to more detailed examination in this study. Of the seven study countries, Thailand has the highest estimated rate of HIV prevalence. The most recent estimate from UN AIDS is that 2.2 percent of Thai adults aged 15-
49 are infected. In India, the percentage HIV-positive is 0.82 percent. Estimates of the prevalence of HIV/AIDS are much lower in other Asian countries examined in this study (Table
13). Although under-reporting is a serious issue, there is little doubt that the epidemic is far more serious in Thailand and other countries of Indochina than elsewhere in Asia.
Table 13. HIV prevalence in selected Asia and Near East Countries.
In the most heavily affected countries of the world, the HIV/AIDS epidemic is having a major impact on mortality conditions. The US Census Bureau estimates that Thailand’s epidemic increased its crude death rate by 16% and reduced life expectancy at birth by 2.3 years, 69.0
49 rather than 71.3, in 1998. Projections to 2010 anticipate similarly large effects in 2010 (US
Bureau of the Census, 1999).
There is a good deal of uncertainty about the course of the HIV/AIDS epidemic and its future impact on mortality in Thailand or in other Asian countries. How rapidly the epidemic spreads will depend on the extent to which individuals curtail or modify high-risk behavior and on the response of institutions to this public health crisis. The impact on mortality and, hence, aging will depend on the development of new forms of treatment that may delay or reduce the progression from HIV to AIDS to death. How these issues will unfold is difficult to foretell. At least in Thailand, current indications are that the response to the epidemic has succeeded in reducing the incidence of HIV/AIDS infection to levels lower than anticipated a few years ago.
Hence, the demographic impact may be somewhat less than incorporated into the UN projections or into projections conducted by other agencies, e.g., the U.S. Census Bureau or the Thai government (the National Economic and Social Development Board). In other countries, the epidemic may emerge with serious effects not anticipated in current demographic assessments.
Uncertainty about the course of the epidemic in Thailand has been explicitly considered by the NESDB working group on HIV/AIDS (Thailand, 1994). They examined three scenarios: a baseline scenario that assumes a continuation of risk behavior at 1993 levels; a medium intervention scenario that assumes a substantial decline in commercial sex and an increase in condom use; and, a high intervention scenario that assumes a very large decline in commercial sex and a very large increase in condom use. The baseline scenario yields cumulative AIDS death of 2.1 million between 1987 and 2020. The medium intervention scenario yields 1.4 million deaths, and the high intervention scenario yields 1.2 million deaths.
50 The impact of AIDS-related mortality on population age structure is very different than the impact of general changes in mortality. Deaths are heavily concentrated among young adults and, to a lesser extent, children. Projections for Thailand, for example, estimate that 76 per cent of all AIDS deaths in 2000 will occur to those between the ages of 20 and 39. Under 10 percent will be to children; under 1 percent will be to the elderly (NESDB Working Group, 1994).
Consequently, the immediate impact of the epidemic is to raise both the child dependency ratio and especially, in percentage terms, the old age dependency ratio. The medium-term impact on the child dependency ratio is muted to the extent that deaths occur among adults who have not yet begun or not yet completed their childbearing.
The impact of the HIV/AIDS epidemic on broad measures of aging is relatively modest in Thailand. Using projections prepared by the Thai government in 1995, the peak in deaths due to AIDS as a percentage of the total population will occur between 2000 and 2005. The projected direct impact of higher mortality during this period is to raise the youth dependency ratio from
0.341 to 0.343 and the old age dependency ratio from 0.103 to 0.104.5 The impact of the epidemic is cumulative in the sense that additional deaths during subsequent periods will lead to further increases in the old age dependency ratio. However, as time proceeds the impact on old age dependency is lessened as cohorts that experienced higher death rates reach old age. This would first begin to occur during the second quarter of the twenty-first century and would partially offset higher deaths among working-age cohorts. Thus, it appears unlikely that the
HIV/AIDS epidemic will lead to a substantially accelerated aging of the Thai population or populations of other Asian countries.
5 Calculations are based on the medium HIV/AIDS intervention variant. These values are calculated assuming that those who died from HIV/AIDS between 2000 and 2005 would not have died from other causes prior to 2005. The impact of differences in the numbers of adults of childbearing age on the number of children is not incorporated.
51 Although HIV/AIDS had a modest impact on the overall burden of population aging, families that experience the loss of one or more adult members will be seriously affected by the epidemic. To a substantial extent, support for the elderly in Asian societies is family-based.
Social norms are changing, but most elderly live with and rely on their children for financial support and personal assistance. With the decline in childbearing to low levels, elderly parents are becoming increasingly dependent on one or two adult children. Because the responsibilities of child to parent are strongly gender-based in many Asian societies, the loss of even a single adult child can threaten the viability of the family support system. We are not aware of estimates of the extent to which AIDS-related mortality will lead to an increase in the number of elderly who have no surviving sons or daughters, but this may be an important emerging issue. In some societies, the epidemic may require the development of and greater reliance on alternatives to family-based support systems for the elderly.
Immigration
International migration is a source of uncertainty for any regional or national population, but for the most part Asia and individual Asian countries have not experienced international immigration in recent decades at levels sufficient to have had a large demographic impact. There are, of course, exceptions to this generalization. Substantial numbers of Filipinos have emigrated to the United States and other receiving countries. The Philippines, Thailand, Pakistan, South
Korea and other Asian countries have exported labor, on a temporary basis, to the Middle East in large numbers. Political upheavals have also generated international migration. The Vietnam
War, revolution in China, partition of India, and the Korean War all generated large scale
52 movements of populations. Such events are difficult, perhaps impossible, to predict and their impact on aging is uncertain.
Data allowing a comprehensive assessment of trend in international migration are limited.
Statistics on international migration flows are rather scarce, but population censuses frequently provide information about the number of foreign born residents. Table 14 presents the number of foreign born or foreign residents enumerated by census or population registers in Asia separately for males and females. In order to provide a comparative picture, migrant populations are also shown for Oceania and North America where migration has played a vital role in population growth.
Table 14. Migrant populations, proportion of foreign born in population, and the growth rate of migrant population; 1965, 1975, 1985 and 1990.
The world’s migrant population rose from 75 million in 1965 to 120 million in 1990, at an average annual rate of 1.9 percent. However, estimates of the number of migrants for intermediate years reveal that the growth rate of migrant stock has not been constant. Instead, it has been rising steadily, rising from 1.2 percent per year during 1965-1975 to 2.2 percent annually during 1975-1985 and reaching 2.6 percent per year during 1985-1990. The trend for
Asia is more mixed. The migrant population of Asia declined at 0.6 percent between 1965 and
1975, increased by 2.7 percent during 1975-1985 and 2.1 percent during 1985-1990. Despite this growth, however, by 1990 international migrants counted for only 2.3 percent of the world’s total population and 1.4 percent of Asia’s. Migration can play a much more important role than
53 has been the cases in Asia. Nine percent of North America’s population is foreign born and 18 percent of Oceania’s. In Asia, perhaps only Singapore has such a large immigrant population.
The limited data on international migration that is available has been used by the United
Nations to make simple projections of net international migration that are incorporated into the population projections used here. Historical data on migration are not available as part of the
World Population Prospects: The 1998 Revision. Projections of international migration rates are summarized in Table 15 for Asia, Oceania, and North America.
Table 15. Net migration, net migration rate, population growth rate, and contribution of net migration to population growth.
The UN anticipates that Asia will experience net out-migration during the first half of the twenty-first century and that Oceania and North America will experience net in-migration. The net number of Asian’s immigrating is projected to be 5.5 million over the period 2000-2005.
This number is projected to decrease until mid 2020s and stabilize at 4 million thereafter. The net migration rate per year (number of net migration divided by population) is relatively small in
Asia, which is projected drop from -0.3 to -0.2 outmigrants per year per 1000 persons. Within
Asia, Southeast Asia is expected to experience the highest net immigration as a percentage of its population (-0.4 per 1,000), whereas countries like Japan, Korea, and Thailand are expected to experience no net migration at all after 2020.
Projected net migration is sufficiently small that it contributes little to population growth in Asia. Its projected population growth is reduced by only 2.4 percent between 2000 and 2005 by net immigration and only by 8.3 percent between 2045 and 2050. In contrast, North
54 America’s projected population growth rate is increased by 41 percent by net immigration between 2000 and 2005 and by 120 percent between 2045 and 2050. Net immigration also plays an important role in Oceania’s overall population growth.
The impact of international migration on the demographic characteristics of both the sending and the receiving populations depend on whether the demographic characteristics of the immigrant population differs substantially from that of the sending or receiving population and also whether immigrant populations are subject to different demographic rates than the sending or the receiving populations. To the extent that immigration is economically motivated, those of working age are more likely to immigrate and particularly those who are relatively young.
However, migrants of working age are often accompanied by family members. Family reunification may lead to immigration by older populations. Schooling opportunities may lead to immigration by younger populations.
Data allowing a comprehensive assessment of the demographic features of international migrants is relatively scarce. However, the US and a few other countries publish some data on the characteristics of immigrants admitted to the US. These data have been used to assess whether immigrants to US are heavily concentrated in a few age groups or not. The values presented in Table 16 are calculated as the ratio of the average number of immigrants in each age group in 1996 and 1997 divided by the sending country’s population in that age group in 1995.
We do not know the extent to which this partial measure of immigration captures the behavior of all immigrants from a country but given the US is an important receiving country for many nations experiencing net out-migration.
There is a clear age-pattern to US immigration. Working age adults, particularly those in their late twenties and thirties, were most likely to immigrate to the US in 1996-97. For the world
55 as a whole, the proportion immigrating was lower for children and for those 40 and older. The elderly were relatively unlikely to immigrate to the US. In general, then, the immediate impact of immigration to the US was to reduce the working age population of the sending countries and to increase the working age population of the US. Note, however, that the immigrant stream was small as compared to the world population and had a negligible impact on the age-distribution of the population living outside of the US. On average, there were 24 immigrants admitted to the
US for every 100,000 persons aged 25 to 34 in the world. As noted above, an immigrant stream of that size was far from negligible from the US perspective and does have implications for the
US age distribution.
Data available for four ANE countries demonstrates the diversity in the age pattern of immigration to the US. In general, the age pattern is not as heavily concentrated at the younger working ages in any of these four countries. In Bangladesh, India, and the Philippines the aged distributions are bi-modal with peaks occurring at older ages. The percentage of elderly
Filipino’s immigrating to the US is particularly high and reflects both family reunification and special rights extended to Filipino’s who served in the US military in World War II. To the extent that the age-profile of emigration from ANE countries is more diffuse, the impact of immigration on the age distribution of the sending countries will be smaller. Again, as noted above, we do not know whether the age-profile of US immigrants is similar or not to net immigration to all countries.
Table 16. Immigrants admitted to US per 1000 persons in the sending country, 1996-97.
56 The overall results suggest that migration patterns by age groups are quite different across countries. One important conclusion is that information about both in-migrants and out- migrants by age group in each country would greatly facilitate efforts to assess the comprehensive effect of immigration on demographic composition. Unfortunately, such information is not available for most countries in the region. Despite the limited information, the influence of net migration on the age distribution of most Asian countries has been quite modest in the past. Unless much larger and persistent migrant streams emerge in the future, Asia’s populations will not be unduly influenced by international migration.
Asia’s economic crisis
During the last few years, several East and Southeast Asian countries have experienced an economic contraction of unprecedented severity. The economic contraction aggravated social vulnerabilities, especially for the poor who are most vulnerable during times of crisis. The health of the older persons, like that of young children and women of childbearing ages, is also particularly vulnerable to the influences of rising medical costs due to inflationary effects of the crisis. There are various social dimensions of Asian Economic crisis (Atinc and Walton, 1998), which include its demographic impact. If the crisis has a sufficiently large demographic impact in the short-run or if the crisis persists for more than a few years, trends in fertility, mortality, migration, and age-structure may be affected in a number of Asian countries.
Table 17 presents economic indicators for the worst affected Asian countries, namely
Indonesia, Korea, Thailand, and the Philippines. The Asian economic crisis translated into rapid contractions in national output and employment. Real gross domestic product declined by as much as 13 percent in Indonesia and as little as 0.5 percent in the Philippines in 1998. Severe
57 exchange rate devaluation created inflationary pressure that resulted in higher prices particularly for imported goods, including drugs and medical supplies. The unemployment rate has also been risen rapidly. In Thailand, it has increased from 0.9 percent in 1997 to 5.3 percent in 1998. In
Korea, it reached an unprecedented 6.8 percent in 1998. Net job losses in the four countries reached over 5 million in 1998. The macroeconomic shock undercut fragile coping mechanisms, and for some may have produced life-threatening declines in income.
Table 17. Macroeconomic indicators of selected Asian countries, 1994-1998
The severity of the Asian financial crisis could have affected, in principle, the demography of countries in a various ways. Mason (1997) enumerates the channels through which economic contraction influences a country’s demography. Fertility can be affected in several different ways, but the most important is probably decisions by couples to postpone marriage and childbearing. In low income settings, a variety of physiological factors, which influence the supply of children, may also respond to economic fluctuations. But these factors probably play a minor role except in cases of widespread famine. Increases in cost of contraceptives may increase unplanned pregnancies. However, many Asian countries such as
Korea, India, and Indonesia are self-sufficient when it comes to contraceptives. Even where the use of modern contraceptives is limited, traditional methods may allow couples to postpone or reduce their childbearing.
Mortality and, more probably, morbidity may increase as a consequence of worsening living conditions. The decrease in income for the lower strata might result in lower food and health care consumption. Deterioration in the quality of publicly provided health care services
58 may have an important impact. The immediate and direct role of the public sector is in the provision of individual and public health services, including, but not limited to, immunization, health education, and primary health services. Public works programs that improve the quality of water supply and sanitation systems also have potentials for influencing mortality. Changes in food policy also may have an important demographic impact (Behrman, 1997). Thus, it is possible that a drop in public investment and expenditure on public services and programs is having an adverse effect on mortality.
The economic crisis may also influence immigration. Labor importing countries may restrict immigration as the demand for labor declines and unemployment rises. On the other hand, workers in labor exporting countries have greater incentives to go abroad. Currency devaluations have substantially increased the wages available abroad in terms of the home country’s currency. Economic contraction at home has reduced wages and increased unemployment. A likely consequence is a rise in illegal migration. What will happen to the number returning is uncertain, because return migration depends on many factors such as wage differentials between countries and alternative employment opportunities between countries. For illegal immigrants the severity of penalties imposed might affect the amount of return-migration.
No clear evidence is currently available to assess the impact of the Asian economic crisis on population growth and structure. Part of the problem can be traced to inadequate data. Current and reliable data on fertility, mortality, and migration are not yet available. Another problem is isolating the effect that is especially attributable to the factors in question. Models are not sufficiently developed to analyze aggregate trends in fertility and mortality and disentangle the impacts of short-term economic phenomena from long-term social and economic change. The third problem is that short-term responses of fertility to economic factors may be quite different
59 from long-term responses. The fertility, mortality, and migration responses can be sometimes temporary and of little demographic significance.
Existing studies of the demographic impact of economic crisis are limited mostly to the
Latin American experience in the 1980s. Some studies have found that fertility is pro-cyclical and mortality is counter-cyclical, but the effect, if any, is temporary. Palloni and Hill (1997) and
Bravo (1997) found that economic fluctuations had substantial contemporaneous effects on fertility and mortality in Latin America. In many cases, however, the effects are not statistically significant, even jointly. Other studies, such as Ortega-Orsona and Reher (1997), reached similar conclusions.
In summary, there is no clear evidence of abrupt and widespread changes in demography responding to economic fluctuations. One important point made by Mason (1997) is that people in many developing countries have shown a remarkable ability to minimize the impact of severe economic crisis. Many developing countries experienced major and prolonged declines in real income, high rates of unemployment, rapid inflation, and significant cuts in social services. In most countries, the demographic impact was not significant. Considering that many East and
Southeast Asian economies are recovering rather rapidly these days, it is not likely that the Asian economic crisis will have a substantial influence on the region’s demography, especially in the long run.
The Implications of Aging and Policy Alternatives
Aging and Economic Growth in Asia
Aging has particularly salient implications for particular sectors (health) or programs
(public pensions), that are discussed in more detail elsewhere in this study. But aging also has broad and important implications for economic growth that transcend these more particular
60 concerns. Important features of the economy are likely to be influenced by population aging, including labor force characteristics, saving rates, poverty and income inequality, interest rates, international capital flows, trade, and international migration (Johnson and Lee, 1987; Deardorff,
1987; Kelley, 1988). In this section, however, we focus only a few important aspects of the connection between aging and economic growth.
Most research on the economic implications of aging focuses on national trends.
However, aging is a regional and global phenomenon. As economies within Asia have become increasingly integrated in the regional and global economy, they are subject to both the demographic changes occurring within their own country, but also the demographic changes occurring elsewhere.
The economic impact of aging has been the subject of more empirically based research in recent years, however analysis relies heavily on models that incorporate our general understanding about how economies work and interact with demographic conditions. This is unavoidable because even the countries furthest along in their demographic transitions have not yet experienced aging to the degree now anticipated. Moreover, aging in Asia has features that will inevitably distinguish it from aging in the West. Aging is occurring more rapidly and, in some instances, at lower levels of income. Thus, the impact of aging in Asia may be very different than elsewhere.
The economic impact of aging will depend to a great extent on the institutional setting and most particularly on the systems that govern the transfer of resources between generations.
These systems include government programs that provide health care benefits and pensions.
They also include family based support systems that are prevalent in Asia. The institutional setting is far from static, however. Family support systems are clearly eroding in many Asian
61 countries. Governments are playing an increasingly important role in determining the responsibility that one generation bears for another. Any assessment of the implications of aging must incorporate the nature of these systems.
During the past few decades, demographic change has favored economic growth in many
Asian countries, and particularly in the successful countries of East Asia. Changes in age structure have led to an increased concentration of populations at the working ages leading, in direct fashion, to rapid labor force growth and more rapid growth in per capita income. Changes in age structure, reduced fertility, and improved mortality conditioned have also had a favorable impact on saving and investment in both physical and human capital. In most Asian countries, demographic conditions will continue to favor economic growth for several decades or more.
With varying timing, however, demographic conditions will turn less favorable to economic growth. The proportion of the population in the working ages will begin to decline relative to the retired population. Saving rates are likely to decline leading to slower accumulation of physical capital. The stock of human capital may also grow more slowly as time proceeds. Of course, demography is not destiny and unforeseen developments in the region may overwhelm these demographic forces.
Slower economic growth should not be viewed with undue alarm where it occurs in countries that have achieved high standards of living and are approaching the end of their demographic transition. Asia’s high rates of economic growth are inherently transitory in nature, a consequence, in part, of rapid demographic transition. Countries that have successfully seized the economic opportunities presented by demographic change are reaching higher standards of living more rapidly than elsewhere. Once higher standards of living are achieved and population aging begins in full-force, economic growth is likely to slow to the more modest levels found in
62 mature economies. It is mistaken to view older populations as economically disadvantaged. They may be relatively poorly endowed in labor resources, but given appropriate policies they may be relatively well-endowed in physical and human capital.
This admittedly optimistic scenario will not play out in countries that fail to harness the potential of an aging population. Policies that encourage the accumulation of physical and human capital, the development of well-functioning financial institutions, a favorable investment environment and integration into the global economy will all become increasingly important.
The challenge for many Asian countries is that the pace of aging is so accelerated and the time to adjust to its new realities so limited.
Population aging influences per capita income directly by influencing the proportion of the population engaged in productive activities and indirectly by influencing the productivity of those who are employed. These two distinct channels can be conveniently distinguished algebraically representing per capita income as the product of output per worker (Y/L) and the proportion of the population in the labor force (L/N)6:
Y/N = Y/L x L/N.
A simple refinement incorporates the fact that "consumption needs" and productivity vary with the age of the individual. Measuring "population" as the number of "equivalent adults" using weights that vary with age and sex provides a refined measure of the standard of living, income per equivalent adult. Likewise, the labor force is measured as the number of "effective workers" using age- and sex-specific weights that capture systematic variation in labor productivity. The
6 Bloom and Williamson, 1998 have recently made use of this identity in their analysis of population and economic growth.
63 equation then states that income per equivalent adult is the product of output per effective worker and the economic support ratio (effective worker per equivalent adult).
The same relationship can be represented in growth terms:
g(Y/N) = g(Y/L) + g(L/N).
The rate of economic growth, as measured by the rate of growth of income per equivalent adult, is the sum of two components: the rate of growth of output per effective worker and the rate of growth of the economic support ratio. The connection between demographic conditions and these two components are examined in the following sections.
Aging and the economic support ratio
Demographic transition throughout the world has systematically produced changes in the economic support ratio that, at times, have accelerated and, at other times, depressed economic growth. In East and Southeast Asia, changes in the economic support ratio have been favorable during the last few decades. Because of population aging, however, the support ratio is expected to decline during this and future decades in East Asia and beginning in 2010 in Southeast Asia.
The countries of South Asia began to enjoy favorable growth in their support ratios only during the 1990s. The benefits will be relatively short-lived there, lasting for only two decades.
During the past fifty years, changes in the support ratio have been less important to economic growth than changes in output per worker. Annual increases tend to be relatively small, but they are quite persistent. Thus, over the course of several decades they have, in the past, contributed to significant increases in per capita income. In the future, changes in the
64 support ratio will depress economic growth. The decline in the support ratio in East Asia is expected to be particularly precipitous, dropping from 0.76 workers per equivalent consumer in
2000 to only 0.57 workers per equivalent consumer in 2050.
The economic support ratio for Southeast Asia, 1950-2050, shown in Figure 7, illustrates how the support ratio typically varies over the transition: declining as the proportion of children in the population increases, rising as the child population stabilizes and the working age population continues growing, then declining as labor force growth slows and the elderly population increases.7 In Southeast Asia, changes in the economic support ratio depressed growth in income per equivalent adult by 0.4 percent (during the 1950s) and contributed as much as 0.25 percent per annum in additional growth in income per equivalent adult in the 1980s. The final panel in Figure 17 summarizes the economic support ratio by identifying the three distinct periods of successive contraction, expansion, and contraction with the average growth rates during each period. This device is used below to compare the experience of Asia's sub-regions and individual countries.
Figure 17. The Economic Support Ratio, Southeast Asia
As a broad generalization, the economic support ratios in each of Asia's major regions and in the individual countries analyzed follow the pattern shown in Figure 17. Figure 18 identifies the extended periods during which the economic support ratio is contributing either to contraction or expansion. To convey the magnitudes of the changes the height of the bars are
7 The figure is constructed using average growth rates during each decade. The support ratio is calculated as the labor force divided by the number of equivalent adults, based on a weight of 0.5 for those under 15 and 1.0 for those 15 and older. Variation in labor productivity by age and sex are not incorporated into these calculations.)
65 drawn in proportion to the magnitude of the changes in each country over the 100-year period.8
Only the Philippines departs from the general pattern shown above. The Philippines did not experience an initial decline in the support ratio. The data presented here do not capture the early stages of Japan's demographic transition. Elsewhere, the cycle in the economic support ratio is present. That said, there are substantial differences among the regions and from one country to the next.
Figure 18. Summary of Economic Support Ratios, Asia and Selected Countries
The values for East Asia are dominated by China. During the last three decades, growth of the support ratio has spurred economic growth by about 0.2 percent per years, but henceforth the support ratio will serve as a substantially dampening force, declining by over 0.5 percent per annum. In South Asia the support ratio grew particularly rapidly during the 1990s and the 2000s, but the impetus provided to economic growth is short-lived as compared with the East and
Southeast Asian experiences. As is true of Southeast Asia, the support ratio will be a negative growth factor beginning in the 2010s. In neither South or Southeast Asia is the support ratio expected to decline nearly as rapidly as in East Asia.
Even more varied patterns become evident as we focus our attention on individual countries. In some countries, of which South Korea and Indonesia are the most notable examples, support ratios have grown more rapidly than elsewhere. In others, India being an example, the support ratio increased very modestly. Likewise, the dampening impact of the support ratio varies considerably among the countries of the region. The average rates of decline
8 The height is proportional to the average of the absolute value of the rates of growth.
66 are greatest in South Korea and Thailand and much smaller in the Philippines, Bangladesh, and
India. Although the correlation is by no means perfect, there is a clear tendency for the countries with the larger gains in the support ratio to experience the greater declines. Compare, for example, the Philippines to South Korea.
The changes in the support ratios are modest as compared with rates of economic growth in the region. In South Korea, for example, the annual increase in the support ratio between 1960 and 1990 was 0.54 percent per annum. Income per equivalent adult grew ten times more rapidly,
6.4 percent, during the same period. In India, the support ratio declined on average by 0.2 percent per annum as compared with an annual increase in income per equivalent adult of 1.8 percent between 1950 and 1990. In a few instances, more dramatic changes in the support ratio occur during a single decade. In South Korea the support ratio increased by 1.3 percent per annum during the 1970s. During the current decade, Egypt and Bangladesh support ratios are projected to increase by 0.6 percent.
In the past changes in the support ratio have tended to reinforce growth in output per worker. South Korea, Thailand, and Indonesia all experienced rapid growth in output per worker and rising support ratios while the less successful economies of Bangladesh and India experienced declining ratios. During the next fifty years, regional differences in economic support ratios may help the countries of Southeast and South Asia to "catch up" with the countries of East Asia. Japan, China, South Korea and other NIEs will experience decline in their support ratios retarding the economic growth of those economies.
67 Aging and output per worker
The impact of declining population growth and aging on growth in output per worker is indirect but potentially more important than the more direct impact of changes in the economic support ratio. Economic research distinguishes two sources of economic growth: technology or increases in factor productivity and factor augmentation or increases in physical and human capital per worker. Demographic factors may operate through both of these channels.
Economists have suggested a variety of reasons why slowing population growth and population aging may lead to slower productivity growth. The most extensive work on productivity and population has examined the response of the agricultural sector to population pressure. Studies by Boserup (1965, 1981), Hayami and Ruttan (1987), and Pingali and
Binswanger (1987) have convincingly demonstrated that innovation in agriculture, induced by population growth, has led to increased productivity of land. These studies do not show that total factor productivity nor that labor productivity increase in response to population growth.
Population growth will accelerate productivity growth in an economy characterized by increasing returns to scale, a feature of recently-developed endogenous growth models (see
Robinson and Srinivasan, 1997 for a recent review.) Under these conditions, slowing population growth in Asia would depress productivity growth. However, empirical studies to date provide little support for the existence of scale economies. Moreover, Singapore and Hong Kong provide persuasive Asian cases of countries relying on trade to realize potential scale economies. Simon
(1981) has argued that a larger population will produce more geniuses, though we know of no direct evidence on this issue. A frequent hypothesis is that an older workforce will be less innovative for a variety of reasons.
68 Countering these hypotheses are the possibilities that diseconomies of scale stifle innovation and that labor shortages may spur innovation that increases labor productivity. One well-known study presents evidence that the slow-down in labor force growth has led to more rapid growth in labor productivity among the industrialized countries (Cutler et al., 1991).
However, no consensus about the connection between productivity growth and demographics has emerged on the basis of the limited evidence available to date.
Aging, saving, and wealth
Under the right conditions, population aging will lead to a substantial increase in wealth or capital per worker. This occurs primarily because: (1) older members of the population have greater wealth, on average, than younger members; and (2) because increased life expectancy leads individuals to save more in anticipation of longer periods of retirement. In countries that are experiencing very rapid demographic transition, the impact on capital accumulation and economic growth may be particularly pronounced.
Public policy may have a profound influence on the connection between aging and the accumulation of wealth. Some retirement systems, e.g., Singapore's Central Provident Fund, institutionalize the connection between aging on wealth. Other systems, such as pay-as-you-go pension programs and traditional family-based systems, may undermine the impact of aging on wealth.9
The emergence of retirement in modern societies requires and is facilitated by the development of institutions that channel resources to the non-working elderly. Initially, the extended family serves this role as elderly parents live with and rely on their children for both
9 This section draws heavily on several recent studies by one of the authors in collaboration with other scholars (Lee, Mason, and Miller, 2000, forthcoming a,b,c).
69 personal and material support. Of course, in many instances the elderly may be serving a productive role within the family, caring for children or providing other valued services. The extended family has persisted throughout Asia and in most countries remains the primary institution that provides support to the elderly.10 But even in Asia, family support systems are being supplanted by new institutions.
Many countries in Europe, the U.S., and Latin America have implemented pension programs that reduce reliance on the family. These systems typically operate on a pay-as-you-go basis, meaning that the retirement needs of those who are currently retired are met by taxing those who are currently working. These programs differ from a family based system in that providing for the elderly is the collective, rather than an individual, responsibility of workers. It is very similar in that both systems rely on transfers from workers to retirees.
Public transfer programs are supplemented by private systems that finance retirement through the accumulation of real wealth. Even in traditional societies, farmers may acquire and improve land or accumulate livestock. In modern societies, we rely extensively on financial markets to purchase stocks, bonds, and other financial instruments, indirectly establishing ownership of real wealth and a claim to the associated future earnings. Individuals may accumulate real wealth acting independently of any organized pension program or they may participate in employee-based pension systems. In a few countries, Chile, Singapore, and
Malaysia being examples, governments mandate and operate funded pension programs either independently or in cooperation with the private sector.
From the perspective of meeting retirement needs, systems based on transfers and systems based on the accumulation of real wealth are in one sense identical. Either system
10 See section III.D.1 on this.
70 produces a stream of income during the retirement years. Likewise, either flow represents wealth to that individual—real wealth in one case, transfer wealth in the other.11 From the perspective of the economy, however, real wealth and transfer wealth are very different. Real wealth is productive and contributes to improved standards of living. Indeed, the rapid accumulation of real wealth is widely viewed as critical to East Asia's rapid economic growth. In contrast, transfer wealth contributes nothing to economic growth.
Although aging and the emergence of retirement have led to an increased demand for wealth, other motives influence saving and forces other than the ones stressed here are also at play. Families may accumulate wealth to protect themselves against short-term fluctuations in their incomes or needs. Some are motivated by a desire to pass on an estate to their children.
Economists differ about the relative importance of these and other saving motives and some question the applicability of the lifecycle model with its emphasis on retirement needs.
Calculating the wealth needed to fund future needs is an exceedingly complex task beyond the ability of even sophisticated consumers. People rely on intuition and rules of thumb and make adjustments as they approach retirement. They rely on advice of varying quality from financial experts. In some instances, mandatory employee-based and public pension programs make the decisions for consumers, institutionalizing the relationship between wealth and demography.
A number of studies have examined the connection between demographics and saving rates (Mason, 1987, 1988; Williamson and Higgins, 1997; Kelley and Schmidt, 1996; Deaton and Paxson, 2000.) They differ in their support for the lifecycle model and in their assessment of the impact of changing age structure. However, these studies are limited because they are based on simpler models that do not capture important features of the dynamic relationship between
11 In other respects the systems may be very different. One may offer a more secure source of funding; rates of return may differ considerably.
71 population and saving and they do not consider the impact of alternative support systems.
Moreover, existing empirical work about the early stages of the demographic transition is much more reliable that estimates about later stages because, to this point, no country has completed the demographic transition.
Simulation techniques provide a means to examine the paths of saving and wealth over the demographic transition if behavior is governed by the life cycle model (Lee, Mason, and
Miller, forthcoming a, b, c). They show that rapidly changing demographic conditions, such as those found in East Asia, lead to equally rapid increases in saving and wealth. As the transition progresses, however, saving rates return to much lower levels while wealth stabilizes at a high level. Given conventional economic growth models, output per worker would also experience a period of rapid, but transitory, growth. Output per worker would eventually grow at modest levels governed by productivity increases alone. Thus, population aging leads to a decline in saving and a slowdown, but not a reversal, in economic growth.
The simulated transition of wealth (Figure 19) and saving (Figure 20) for Taiwan are compared to simulated changes given a much slower demographic transition such as experienced in France or the United States. While the total fertility rate declined from 6 births per woman to 2 births per woman in only 30 years in Taiwan, the fertility transition in United States took approximately 100 years and in France occurred over a period of two centuries. Life expectancy also increased much more rapidly in Taiwan than in the West.12 In each of the three simulations, demographics are identical at the end of the transition; they differ only in their timing and speed of the transition.
12 These simulations do not include the impact of the baby-boom. See Lee, Mason, and Miller, 2000 for US simulations based on US economic and demographic data.
72 Prior to the demographic transition, the demand for wealth was relatively small, equal to or less than twice gross domestic product (GDP). At the end of the demographic transition, the life cycle demand for wealth exceeds 5 times GDP. In each of the three scenarios, the demand for wealth experiences a period of acceleration followed by convergence to the high post- transition level. In the simulation based on Taiwan demographics, however, wealth grows much more rapidly and saving rates reach a much higher peak.
Figure 19. Wealth/Output: Taiwan, US, & France.
Figure 20. Saving Rate: Taiwan, US, & France, 1800-2100.
The simulations presented here do not distinguish real wealth from transfer wealth. The path of real wealth and real saving will depend on the extent to which life cycle needs are being met by transfer systems. In the United States, for example, Social Security (OASI) wealth is 1.75 times GDP or about 15 trillion dollars for the year 2000. This amounts to almost half of the total demand for life cycle wealth in the year 2000. Thus, demographic change in the U.S. should have a more attenuated impact on saving and wealth in the absence of substantial reform to the
Social Security system (Lee, Mason, and Miller, 2000). In Egypt and most Asian countries, public transfers are much less important but family transfers are pervasive. Until recently in
Taiwan, for example, the elderly relied almost entirely on transfer wealth obtained by living with their adult children. If the family support system were to persist in Asia, then changing demographic conditions would not lead to high saving rates and rapid accumulation of real
73 wealth. Rather, it is demographic changes in conjunction with the erosion of the family support system that lead to high rates of saving and real wealth (Lee, Mason, and Miller, 2000.)
Changes in saving and investment throughout the world appear to be broadly consistent with the demographic forces described here. Data available for 104 countries are used to compare investment rates in 1960 and 1990 in Figure 21.13 Many countries are clustered around the 45-degree line indicating similar investment rates in 1960 and 1990. Countries with high rates of investment in 1960 tended to have lower rates of investment in 1990. These declines occurred mostly among Western countries that were relatively advanced in their demographic transitions. Fourteen of the twenty-two countries with investment rates of twenty-five percent of
GDP or more in 1960 were Western. As a group they experienced a decline in their investment rate by five percentage points by 1990.
The experience within Asia is broadly supportive of the thesis that rapid demographic transition leads initially to high rates of saving and investment. Asian countries that experienced rapid demographic transition experienced higher investment rates (and higher saving rates) in
1990 than in 1960. Some of the countries in question had very low investment rates in 1960 but among the highest investment rates in the world by 1990. South Korea and Indonesia are excellent examples. The experience of the slow transition countries was quite different. The
Philippines, India, Pakistan, Bangladesh, and Egypt, with delayed and relatively slow transitions, did not experience a substantial increase in their investment rates.
Figure 21. Investment rates, 1960 versus 1990, countries of the world.
13 Investment data are used here rather than saving data because of their greater availability. The two variables are high correlated.
74 Among the countries of Asia, only Japan is sufficiently far along in its demographic transition to experience a decline in its saving ratio because of demographic forces. Many scholars, though not all, believe that saving rates will decline precipitously as a consequence of population aging. Recent forecasts anticipate a decline in gross national saving rates to between about 5 and 15 percent of GDP as compared with a peak of 40 percent achieved during the 1970s
(Figure 22).14
Figure 22. Gross national saving, Japan 1885-1997 and forecasts.
The evidence about population, wealth, and saving is suggestive rather than conclusive. It is clear that substantially increased wealth and a period of high saving rates are necessary if the elderly are to support themselves during retirement in a self-reliant fashion rather than by depending on transfers from younger generations. Aging will not necessarily lead to a substantial increase in wealth, because other factors may undermine the demand for wealth in an aging society.
Concluding observations on aging and economic growth
Unforeseen developments may intervene in any long-range assessment of the economy.
Perhaps technological breakthroughs will lead to productivity growth that is much more rapid than has been true of the past. The strength of the US economy during the last few years provides fuel for optimistic assessments. Demographic change, however, will almost surely dampen economic growth in many countries during the coming decades. Two trends are particularly
14 See Mason and Ogawa, forthcoming for a summary of recent forecasts of Japanese saving and a more detailed discussion of the issues.
75 salient. The first is the decline in the economic support ratio. The number of consumers will grow more rapidly than the number of workers. The second is the impact of aging on saving and investment. After a period of extraordinarily high rates of saving and investment in many Asian countries, an era with lower rates of saving and investment and slower growth in labor productivity appears likely.
Understanding the impact of demographics on the macro economy is important for at least two reasons. First, long-range planning depends on a realistic assessment of long-range economic prospects. Unduly optimistic forecasts of long-term economic growth leads to irresponsible fiscal policy and the continuation of unsustainable pension and health care programs. Second, a better understanding of the implications of aging for the economy helps to avoid the implementation of policies designed to pursue unrealistic and inappropriate objectives.
One simple example of this is the frequent call for higher national saving rates for economies with aging populations. As discussed above, the saving rate needed to meet the retirement needs of an aging population depends to a great extent on the nature of the demographic transition.
Given a rapid transition, high saving rates are required for a limited period. Later in the transition, however, retirement needs can be satisfied with a relatively modest national saving rate.
Understanding the connection between population aging and the macro economy is also important in framing a wide range of policies appropriate to aging societies. This is an issue that we take up in our concluding section, below.
76 Health and Health-Care Policy
Population aging is expected to place an increasing burden on the health sectors in Asia.
Increases in the levels of economic development, income, education, housing, sanitation and health care delivery can be expected to lead to reduced mortality, increased longevity and increased demand for health services by the aging population. Among the largest and most advanced OECD countries, including Japan, health care spending for the elderly (age 65 years and older) is roughly 4 times the per capita rate for those under age 65 accounting for some 2.5-
5% of GDP (Anderson and Hussey 2000, p. 195). This portends of a substantial shift in resources as the Asian countries seek to treat acute and chronic conditions of a non-communicable nature, which will afflict the aging populations of these countries. Aging populations in Asian countries can expect to suffer from cardiovascular diseases (CVD), cancer, (particularly lung cancer), chronic obstructive pulmonary disease (COPD), musculoskeletal conditions (including osteoporosis), dementia, and blindness and lesser visual impairments (WHO 1998, pp. 106-110).
Simultaneously, however, infectious diseases, both new and resurgent, will continue to plague
ANE countries (WHO 1999, Chapter 2, pp. 13-27). In this regard, India is a leading example where tuberculosis threatens all ages and the elderly are particularly vulnerable to infection.
Therefore, it is extremely important that countries pursue an efficient and effective health policy as treatment of communicable and non-communicable diseases place claims on resources.
National health expenditures
An important health policy consideration is the allocation of resources to the health sector from the overall production of the economy. This is typically measured as the percentage share of national health expenditures within GDP. However, for data on health expenditures to be comparable across counties, the national health accounts must use standardized definitions and
77 methods. One of the best sources for comparable data on national health expenditures is the
OECD (OECD 1999). Unfortunately, the standards of accounting are not uniform in most developing countries. Using World Bank data, Jowett (1999) reports that national health expenditures, as a percent of GDP, actually fell in a number of low-income countries, including
Bangladesh and Nepal, during the period 1990 and 1995. It is difficult to determine whether this is due to general social-economic conditions and shifts in policy or is a remnant of inconsistent accounting. There has been a movement to improve national health accounts in many developing countries, including the Philippines, Egypt, India and Thailand (Berman 1997; Rannan-Eliya, et al. 1997; Tangharoensathien, et al. 1999). Differing methods can account for large differences in estimated spending. For example, Tangharoensathien, et al. (1999) compute health care consumption expenditures using a National Health Accounts prototype to be 3.00% of GDP in
Thailand 1994, while the National Economic and Social Development Board (NESDB) method renders 5.01% for the same year. Using WHO data, national health expenditures for the ANE countries are reported in Table 18.
Table 18. National health expenditures as percentage of GDP, 1997, WHO estimates.
The estimates for Bangladesh and the Philippines are quite different than World Bank estimates reported by Bos, et al. (1998) a few years prior (see Table 19). In addition, the Korean estimates of national health expenditures as a share of GDP vary a great deal between the WHO definition and the OECD definitions, 6.7% vs. 6.0%.
Table 19. National health expenditures as percentage of GDP, World Bank estimates.
78 According to the OECD, national health expenditures as a percentage of GDP were 7.2% in
Japan and 6.0% in Korea in 1997. In the case of Japan, national health expenditures increased from 3.0% to 7.2% of GDP during the period 1960-97. In Korea, health expenditures increased from 2.3% to 6.0% of GDP during the period 1970-97 (See Figure 23).
Figure 23. Rates of total expenditure on health by GDP.
These percentages are approximately half of the US expenditure rate of 13.9%. Nonetheless, health expenditures as a percent of GDP have increased for all the majored developed countries during the period 1960-97 (See Table 20). Aging populations are surely a contributing factor to the rise of health expenditures in OECD countries. There is some debate, however, the effect aging has on health expenditures.
Table 20. National health expenditure as percentage of GDP 1960-1996, G7 countries, Australia and South Korea.
Aging and health expenditures
If aging increases the demand for medical care, then the changing age structure will drive-up per-capita health care expenditures and the size of the health sector. Analyzing OECD data, O’Connell (1996) finds that for some countries, including the United States, Japan, Italy,
Austria, Canada, Finland, and Greece, age structure has a significant effect on per capita health care expenditures. Some previous studies had often failed to find age as a significant factor, ceteris paribus, in determining health expenditures (Gerdtham, Søgaard, Andersson, and Jönsson
1992; Getzen 1992). Utilizing an aggregate US time-series 1960-87, Murthy and Ukpolo (1994)
79 do find that the age structure is an important determinant of health care spending. Similar results have been found with OECD cross-sections. Overall, however, the results are mixed.
All these studies rely on aggregate country data. Zweifel et al. (1999) utilize micro data on health expenditures by elderly individuals in Switzerland spanning the period 1983-1992.
They find no effect of aging on health care expenditures for individuals who die at age 65+.
Their results indicate that the main determinant of spending is the years remaining till death, as very substantial health care expenditures occur in the last two years of life–an observation consistent with US Medicare data (HCFA 1998). In the US in 1996, Medicare beneficiaries in the terminal year of life comprised 6.4% of all beneficiaries but accounted for 20.5% of program payments (HCFA 1998, p.42). Average expenditures for beneficiaries who died were 3.8 times higher than for those who survived (HCFA 1998, pp. 42-43). In the Zweifel et al. (1999) study, there is little difference, for example, between the expenditures for someone who dies at 70 and someone who dies at age 80. The main driver of individual health expenditures is death not age.
Populations which age due to increased longevity will push health expenditures into higher ages.
Although individual health care expenditures may not depend on age, aggregate spending will be greater for older populations because the proportion of the population near the last few years of life will be greater. If we look at projected mortality as a ratio of population 15-64 with this view in mind, Japan and Korea can expect rising health care expenditures and a rising burden (See
Figure 24).
Figure 24. Burden of death, India, Japan and South Korea.
A similar conclusion can be drawn with respect to long-term care expenditures associated with severe disability. The prevalence of severe disability is falling in many OECD countries
80 among the younger old, age 65-75. Nevertheless, aggregate expenditures for long-term care, including institutional care and home care, are projected to rise as a percentage of GDP because the proportion elderly in the population will be larger (Jacobzone, Chambois, Chaplain and
Robine 1998).
Income and health expenditures
The expansion of a country’s health care system is thought to accompany general economic development and rising per capita income. Evidence of increasing demand for health care services is captured in an estimated income-elasticity, which indicates the sensitivity of the health care demand to changes in income. An income-elasticity is a unit-free measure of this responsiveness and is calculated as the percentage change in the quantity demanded of health care goods and services divided by the percentage change in income. An income-elasticity equal to unity implies the health care system will grow proportionally with income and, thus, glean a constant share of society resources. An income-elasticity greater than unity implies the health care system will grow more than proportionally with income and absorb an increasing share of national production. Finally, an income-elasticity less than unity implies the health care system will grow less than proportionally with income and claim a decreasing share of resources and output.
Table 21 summarizes the findings for a large number of studies, which estimate income elasticities for health care from expenditure data. Using aggregate data, estimates for OECD countries elasticities are around 1 or greater than 1. In developed countries, most health care consumers are covered by health insurance. Given that a household is covered by health insurance, the purchasing is achieved through the insurance pool rather than household income.
81 Therefore, ceteris paribus, household income is an unimportant determinant of health care demand in insured societies and income elasticities are near zero when estimated on micro data
(Getzen 2000). Thus, the difference between estimates based on aggregate country data and household data can be explained by the presence of insurance. This hypothesis is consistent with data from Southeast Asia. If we estimate health care spending from household expenditure surveys in Indonesia, the Philippines and Thailand in 1980’s, a time when health insurance was extremely limited in these countries, estimated income-elasticities are significantly greater than one. This implies that health care spending will grow in these countries with the rise in economic development independent of aging.
Table 21. Estimated income elasticities form health care expenditure data.
Health care financing systems
To care for their populations, health care systems must finance their use of resources.
This brings major issues of equity and efficiency in the finance and delivery of care. Countries must decide what type of health care to produce and for whom and do so efficiently. Many developing countries, for example Egypt, India, Indonesia, and the Philippines, focus on maternal and child health because it is viewed as most “cost-effective,” in terms of lives saved.
The financing of long-term care in these countries is a low priority, because providing primary care to all has yet to be achieved and aging has not become a major issue. In the case of India efficiency dictates infectious and communicable disease control, as diseases like TB are extremely prevalent (Table 22). Different choices face Japan, Korea and Taiwan. Japan, like the
U.S., is choosing to heavily subsidize the health care for the elderly. Japan is also heavily subsidizing long-term insurance. As there is very little economic justification in terms of
82 efficiency (i.e., public good and externality arguments) for these subsidies, only equity arguments prevail. Political economy would imply that the elderly are claiming an increasing share of resources because they are an increasing portion of the electorate in the U.S. and Japan.
Table 22. Tuberculosis among populations 60 and older, 1990.
Egypt
Egypt is embarking on major reform of the health system. A largely fragmented system will be re-organized with the Ministry of Health and Population (MOHP) responsible for delivery and the Health Insurance Organization (HIO) responsible for financing (Partnerships for
Health Reform 1999). Heretofore, each organization has engaged in both financing and delivery of services with HIO operating over 300 health care facilities, including hospitals, clinics and dialysis units (Nandakumar, et al. 2000). The separation of financing and delivery is hoped to encourage greater efficiency in production and consumption of services while improving quality and equity. Due to limited resources, the new health system model will focus on primary health care with the goal of providing a basic package of services for the entire population (Partnerships for Health Reform 1999). The vanguard project of Egypt’s health reform is the health insurance program for school children. Under Law 99, passed by the People’s Assembly of Egypt in June
1992, all school children are provided health insurance through HIO. This represents a landmark event in health care financing in Egypt as the number of person covered increased from 3.75 million in 1988 to 14 million in 1993 as a result of Law 99 (Nandakumar, et al. 2000). This has increased the percentage of the population covered from 5 to 25% over this period (Nandakumar, et al. 2000). Prior to the introduction of Egypt’s School Health Insurance Program (SHIP), HIO primarily insured government employees. Coverage was mandated and financed through a wage-
83 based premium between 1-4% depending on the employees’ salary and modest cash co- payments. The premium expense is shared between the employer and the employee with the employee’s share not to exceed 1% of wages. The financing of the SHIP system is accomplished through general tax revenues, a dedicated cigarette tax, a mandated household premium and 33% co-payment on drugs (Nandakumar, et al. 2000).
Despite recent advances, health insurance coverage for the elderly in Egypt is extremely limited. Approximately 2% of the elderly population is covered through the Health Insurance
Organization (HIO) by virtue of being retired government employees, spouses or widows thereof
(Nadakumar, El-Adawy and Cohen 1998). There are only a small number of nursing home facilities and generally very few facilities and services specifically designed for and provided to the elderly. Although care for the aged is seen as an important policy goal (Partnerships for
Health Reform 1999), the constraint on resources has forced it into a low priority position as
Egypt reforms its finance and delivery system.
India
The Indian health care system includes a wide variety of traditional medicine delivered along side a government operated Western-style hospital system. The private sector dominates both the finance and delivery of health care in India. Seventy-five percent of all financing comes from households’ out-of-pocket payments and much of their spending is directed toward private out-patient services (Berman 1998).
84 Japan
Japan has a system of universal coverage achieved through a patchwork of more than
5000 independent insurance plans, which are generally employment-based and funded through premium contributions from employee/beneficiaries, employers (including the public sector) and government tax revenues (Ikegami and Campbell 1999; Arai and Ikegami 1998). There is no private health insurance and the entire population is covered through these public schemes
(Oliver, Ikegami and Ikeda, 1997). There are three basic types of plans. The first type is based on large employers in the private sector through Society-Managed Health Insurance (SMHI) plans and the public sector through Mutual Aid Associations (MAAs), with premiums roughly split between employer and employee. The second type is Government Managed Health Insurance
(GMHI), which is a national scheme that covers employees of small enterprises and their dependents. The GMHI plan is subsidized at a 14% rate by government tax revenues. Under these employment-based plans additional financing is garnered through co-payments, which generally run at 10% for employees and 20% (inpatient) and 30% (outpatient) for their dependents (Arai and Ikegami 1998; Ikegami 1991). The third type, Citizens’ Health Insurance
(CHI), covers the self-employed and retirees and is based within local communities through the city, town and village governments (Ikegami and Campbell 1999, pp. 57-58). Government subsidies for CHI plans average 50%. In addition to direct government subsidies, the premium structures in all three types of plans are designed to generate cross-subsidies from the rich to the poor and young to the old. In other words, the elderly pay less than actuarially fair rates even on a tax-adjusted basis.
In addition the basic health insurance described above, Japan has instituted a mandatory long-term care insurance program (Kaigo Hoken) for the elderly effective 1 April 2000
85 (Campbell and Ikegami 2000; Ikegami 1999; Ikegami 1998; Ikegami 1997; Watts 2000). This mandatory public program requires everyone age 40 years and older to contribute premium payments to the national-insurance pool. For workers aged 40-64 the program will provide services in the event of disability. Their mandatory premium payments can be viewed as a wage tax and, in this regard, is similar to that which funds the US Medicare program. In addition, general tax revenues will fund 50% of the program with this burden shared by the national and local governments (25% national, 12.5% prefectures and 12.5% municipalities). The beneficiaries, mostly frail elderly, will pay a 10% co-payment at the point of service for nursing care (Koinuma 1999). A modest monthly premium, expected to average ¥2800 ($25) in the first year of the program, will be deducted from the public pension payments of those aged 65 years and older (Creighton and Ikegami 2000). Financed services consist of institutional care and community-based care (including home care) while medical services will continue to be financed through the existing health insurance system. The system will use a managed care approach with pre-approval necessary for services to be covered. Service levels will be defined as a financial amount depending on the medically determined health needs of the beneficiary and will range from ¥61,500 to ¥358,300 ($567 to $3305) per month (Creighton and Ikegami 2000). The degree of family support will play no direct role in determining benefit levels (Ikegami 1998). Providers of nursing-home care and home-care services will be local public providers as well as private sector competitors. In principle, the new scheme will also permit international providers to compete and receive payments with some US firms planning joint ventures to enter the Japan elder care market (Saphir 2000).
86 Philippines
There two major health policy developments in the Philippines. The first is the plan for national health insurance (Busse and Schwartz 1997) and the second is the decentralization of public health care delivery system (Bossert, Beauvais and Bowser 2000).
The Republic of the Philippines is attempting to achieve universal health insurance coverage by the year 2010. The National Health Insurance Act of 1995 (Republic Act No. 7875) instituted The National Health Insurance Program and established the Philippine Health
Insurance Corporation (PhilHealth) for this purpose. The National Health Insurance Program
(NHIP), formerly known as Medicare, is a health insurance program for Social Security System
(SSS) members and their dependents and is based on the standard principle of risk-pooling with additional cross-subsidization from the healthy to the ill and from the relatively high-income to low-income. PhilHealth is now the legislatively mandated administrator of the Medicare program. As a result, the SSS, which administered the Medicare program for private sector workers, has transferred the administration of the program to PhilHealth.
The main focus is coverage for hospitalization. The current plan covers hospital charges
(i.e., room and board, drugs, and diagnostic tests), professional fees, surgical expenses, and surgical family planning services, including vasectomy and tubal-ligation. The program is quite modest but is expected to evolve over a fifteen-year period to a comprehensive system in both the level of benefits and proportion of the population covered. Five years into the program few of the early goals have been met and the program is not likely to reach its goal of complete and comprehensive coverage by the year 2010.
Under “Devolution,” control and financing of the public health care delivery system has been transferred to local government units. The National government provides some funding by
87 “revenue-sharing. ” Provincial governments are expected to collect taxes to finance their contribution to the health care delivery system. Tax collection administration and enforcement is weak in many areas of the Philippines making it difficult for local governments to finance their share. Eventually it is hoped that public sector providers will become more efficient as they compete with the private sector for PhilHealth payments and in the process become more self- sufficient. In this regard, the Philippines is pursuing a managed competition approach in combination with a single-payer system (i.e., PhilHealth).
South Korea
The South Korean health care system is financed through employment-based insurance schemes. Participation is compulsory and coverage is universal since 1989 (Yang 1996, Shin
1998). Employees of the Government and private schools are covered under a single insurance pool, the Korean Medical Insurance Corporation. Industrial employees are pooled onto Medical
Insurance Societies, which number approximately 150. Together, these wage earners and their dependents constitute approximately 47% of the population (Shin 1998). Participation and premiums are compulsory with the legal burden of the premium roughly a 50-50 split between the employee and employer. Premium rates range from 3.2% to 3.8% of wages (Shin 1998). The self-employed and their dependents in urban and rural areas are grouped into Medical Insurance
Societies of which there are several hundred covering approximately 49% of the population
(Shin 1998). These insurance pools are funded by premium contributions from the beneficiaries determined by income and wealth tests and by government subsidies. However, the degree of cross-subsidization is relatively modest compared to social insurance schemes in most OECD countries. The poor are covered by Medical Aid program, with approximately 4% of the
88 population eligible (Shin 1998). Although the proportion of the population covered is complete, the coverage level and services included in the benefit package are quite limited. As a result approximately 50 % of health care expenditures are financed through cash payments by the beneficiaries at the point of service. These so-called “out-of-pocket” payments include co- insurance and deductibles for covered services and full payments for uncovered services. For covered outpatient services a patient pays a flat fee/deducible plus a co-insurance rate. The outpatient co-insurance rates are 30% for services rendered in outpatient clinics, 50% for hospital outpatient services and 55% for general hospital outpatient services (Yang 1996) and are obviously designed to discourage hospital outpatient use. The co-insurance rate for inpatient care is a uniform 20% (Yang 1996). These cash payments represent a significant portion of household spending for the low-income and the elderly. Lifting this burden remains a major challenge for the Korean health care system (Table 23).
Table 23. Out-of-pocket payments as percentage of total health expenditure: South Korea
Health care providers are reimbursed on a fee-for-service basis under government regulated fee schedules. Korea is, however, reforming the payment system with a USA style case-based approach using diagnostic related groups (DRGs) for hospital payments (Fetter 1991;
Averill 1991) and resource-based relative value scales (RBRVS) to construct physician payments
(Hsiao, 1988a, 1988b, Hsiao and Dunn 1991, Hadley 1991). These payment reforms are expected in increase efficiency and curtail the rapid growth of expenditures, which seem unabated under fee-for-service despite hefty out-of-pocket payments.
89 Thailand
Thailand’s public sector health delivery system consist of a nationwide system of community health service stations (village health stations), health centers, community (district) hospitals and provincial (regional and general) hospitals with provincial hospitals providing extensive outpatient (ambulatory care) and inpatient services. This system is administered by the
Ministry of Public Health and is the main provider of services in rural Thailand. The relatively high-income urban areas support a vibrant and growing private health sector with numerous private hospitals and clinics and physicians in private practice. In addition, many provincial hospital-based physicians and other physicians, who are public-sector employees, maintain private practices. In the Bangkok Metropolis nearly 50% of the physicians and dentist are employed in the private sector and account for approximately one-half of the provision of services associated with out-patient visits (Mongkolsmai 1997). Forty-two percent of the hospital beds in Bangkok are in the private sector compared to 16% in other provinces (Mongkolsmai
1997). Inpatient admissions in the public and private sectors are in roughly the same proportion as the supply of beds in the respective sectors.
Approximately 59% of the Thai population is covered by some type of health insurance
(Mongkolsmai 1997). There are four main categories of publicly sponsored insurance schemes
(Mongkolsmai 1997; Tangcharoensathien, Supachutikul and Lertiendumrong 1999;
Supakankunti 2000). The first is public assistance for the poor, elderly (aged 60 plus years) and primary school children, which provides free medical care through the public delivery system and is financed though general tax revenue. This scheme is means-tested for low-income households. The beneficiaries are given a health card, which gives them free access to the (sub- district) health centers and community (district) hospitals. This scheme uses “
90 limit access to higher-level facilities, as written referral is required. In 1993 36% of the population including 3.5 million elderly persons were covered by this scheme (Mongkolsmai
1997). The second major category of public insurance is the Civil Servant and State Enterprises
Medical Benefit Scheme, which is financed 100% through government tax revenues. The public sector as employer provides this generous fringe benefit to employees, their spouses, parents and up to three children under age 18 years thus covering 11% of the population (Mongkolsmai
1997; Tangcharoensathien, Supachutikul and Lertiendumrong 1999). Government retirees receiving pensions carry this benefit with them into old age. The third major category is the compulsory Social Security scheme, which covers private sector employees in firms with more than 10 employees. Social Security is financed through premiums based on 4.5 % of wages with contributions split evenly between three parties: employer 1.5%, employee 1.5%, and government (tax revenues) 1.5% (Mongkolsmai 1997; Tangcharoensathien, Supachutikul, and
Lertiendumrong 1999). The premiums, therefore, consist of a 3% wage tax plus a 1.5% subsidy.
This generates within Social Security a substantial cross-subsidy from high-wage earners to low- wage earners with a maximum premium differential of 3:1. Employers with 10 or more employees are required to either participate in Social Security or provide a more generous private scheme. Approximately 8% of the population is covered by Social Security and Workers’
Compensation, which is for on-the-job injury (Mongkolsmai 1997).
The fourth type of public insurance scheme is the Health Card Program, which is a voluntary program mainly targeting near poor and middle-income rural households. The annual premium is 1000 baht (approximately $25) per household up to a maximum of 5 persons with households contributing 50% and government tax revenues financing 50% (Supakankunti 2000).
The Health Care Program (HCP) gives access to the public delivery system in much the same
91 way as the welfare scheme for the poor, elderly and school children, as the cardholders are restricted to MOPH providers. Because it is a voluntary system, the HCP is subject to adverse selection (Supakankunti 2000) a problem not faced by the compulsory Social Security scheme.
In addition to these four publicly sponsored insurance schemes about 1-2% of the population is covered by private health insurance (Mongkolsmai 1997; Supakankunti 2000).
Policy issues in health care financing design
Health insurance can spur substantial expansion of the health sector by providing consumers of services with strong, effective demand for health care, which thereby generates a ready source of revenues for providers. A health insurance system, however, must finance payouts with some combination of taxes and insurance premiums.
Risk-Sharing
Health insurance provides protection against the risk of medical expenses by allowing the insured to share that risk with other members of the group. In any particular period, insured individuals faced with adverse health outcomes and associated medical expenses will receive payment either directly or indirectly from a common financial pool. That financial pool is supported by all members of the group through taxes or premium contributions. Because all contribute but few claim benefits, the pool remains financially solvent. With subgroups placing claims on the group's financial resources in a more or less random fashion, the risk is spread over many individuals and is substantially lower than that faced by an individual in financial isolation.
Although most public insurance and some private insurance perform a redistributive function as well, risk-sharing is considered the fundamental function of insurance. Private insurance
92 markets, however, may fail to spread risks properly owing to informational imperfections. This has important implications for health policy.
Informational imperfections
Asymmetry in information about the actions of insured agents and the potential risks faced by insurers leads to partial failure of insurance markets. This insurance problem falls into two categories, moral hazard (hidden action) and adverse selection (hidden risk) (Pauly 1968,
1974; Rothschild and Stiglitz 1976).
Moral hazard. Moral hazard is most broadly defined as an informational problem of hidden action and unobservable states of nature (Leu 1986; Arnott and Stiglitz 1988). In the case of medical insurance, moral hazard arise whenever an individual can affect the probability of illness or accident (Pauly 1974; Marshall 1976; Arnott and Stiglitz 1988, 1990), or the level of utilization or expenditure, given a particular outcome (Pauly 1968; Zeckhauser 1970). The insurer cannot observe the actions of the insured. Therefore, premiums and associated incentives for prevention are not at appropriate levels to induce optimal behavior, and competitive equilibrium is inefficient (Arnott and Stiglitz 1990). As a result of moral hazard, insurance claims will increase more than proportionally with the increased level of insurance coverage.
Some risks, therefore, are uninsurable. In sum, moral hazard leads to less than full insurance coverage and higher insurance and medical costs. If public-sector financing increases insurance coverage and benefits, expenditures will likely rise because of moral hazard (Leu 1986).
Co-payments limit the moral hazard effect by leaving the insured with some risk and direct cost (Pauly 1968, 1974) and thus an incentive to restrain expenditures (Zeckhauser 1970).
In the case of traditional reimbursement insurance, two forms of co-payment are typically used, deductibles and co-insurance. The effectiveness of co-payments in restraining moral hazard may
93 be reduced, however, by supplemental insurance coverage (Pauly 1974). A sound long-term care insurance program would require some level of co-payment.
Deductibles leave some initial level of expenditure completely uninsured. The insured individual must reach a particular level of out-of-pocket expenditure (the deductible amount) before the insurance coverage becomes effective. The deductible leaves the individual uninsured for small expenses. This has the effect of reducing administrative costs because some small expenses never enter the system. Moreover, deductibles discourage the use of formal medical services for trivial conditions. With the prudent use of deductibles, the risk-sharing function is kept largely intact as large and catastrophic expenses continue to be insured. Sound actuarial management of any insurance scheme calls for substantial deductibles for many types of covered services. The failure of developed countries to implement relatively large deductibles under social insurance has contributed to the escalating costs of health care within those societies. It is important to incorporate deductibles and other forms of cost sharing into a health insurance scheme from the outset because the political economy of health care makes it difficult to remove generous entitlements once they are granted.
Under co-insurance, a percentage or per-unit amount of expenses is borne by the insured at the point of service. User fees for medical services fall under this general rubric. These direct out-of-pocket expenses discourage excessive use and thus reduce moral hazard. The risk-sharing function, however, is partially sacrificed under co-insurance. Therefore, co-insurance rates should not be so excessive as to render the insurance coverage ineffective. Under the usual assumptions associated with individual preferences toward risk, it is never optimal to impose a
100% co-insurance (0% insurance) rate (Shavell 1979). What is needed is a balance between economic incentives and risk-spreading (Zeckhauser 1970).
94 In developed countries, co-insurance rates of 10-20% are common. For example, under
Japan's employer-based system, which covers approximately 63% of the population, employees face a 10% co-insurance rate while their dependents face rates of 20% in-patient care and 30% for out-patient care (Ikegami 1991, p. 91). The appropriate co-insurance rates vary among countries and service types, depending upon the responsiveness of supply and demand to rate variations. In the case of Korea high deductibles and co-insurance rates lead to out-pocket- payments of approximately 50%. This substantially limits moral hazard, but reduces risk- spreading by half.
Given the presence of moral hazard, public and private insurance plans that are efficient will provide less than complete insurance. Under free-market systems, this situation presents an opportunity for a secondary insurer to provide supplemental coverage to the purchaser or beneficiary of the primary plan. Supplemental insurance typically parallels a primary insurance contract or benefit plan and provides financial payment for the uncovered portion. In other words, supplemental coverage usually insures for the deductibles and co-insurance specified by the primary plan. Supplemental insurers rarely provide benefits or cover services that are not also partially covered by the primary insurer with good reason. Supplemental coverage, by providing increased insurance, induces additional moral hazard. The cost of the increased moral hazard, however, is borne largely by the primary insurer and is reflected in higher premiums or taxes there. Thus, the premium on supplemental coverage is low relative to the cost of moral hazard precisely because the primary insurer is bearing some of that cost. In relation to the marginal social cost, the supplemental premium is too low. Overall, supplemental coverage leads to overinsurance, resulting in increased moral hazard, which in turn renders overutilization. For this
95 reason, some analysts call for taxation or prohibition of supplemental coverage unless that coverage is provided by the primary insurer.
Adverse selection. Adverse selection, a classic asymmetric information problem, results from unobservability of the risks by the insurer. It occurs when high-risk patient-consumers self- select favorable insurance coverage, thereby driving up insurance premiums and driving out low- risk patient-consumers, causing in turn a further increase in premiums. Eventually the insurance market shrinks, leaving some individuals and households uninsured or with less than full coverage (Rothschild and Stiglitz 1976). If private insurers can screen out high-risk patients, those patients may be uninsured. Moreover, screening of patients uses valuable resources and may result in a misallocation (Crocker and Snow 1986; Borenstein 1989). In either case, some segment of the consumer market is left uninsured or underinsured.
Universal coverage
With universal coverage, all members of a society are covered by health insurance. This is typically accomplished by requiring mandatory participation in a public health insurance scheme, requiring everyone to purchase private insurance, or a combination of the two. The main advantage of universal coverage is that it eliminates the adverse selection problem by preventing individuals from self-selecting and insurers from screening. Nevertheless, universal coverage accentuates the moral hazard problem. Thus, total expenditures will rise unless cost-containment measures are introduced. In this regard, it is particularly important how payments to private providers are structured.
96 Provider payment mechanisms and policy
Several payment mechanisms exist for the reimbursement of health care providers.
Among these are fee-for-service, capitation, and prospective payment systems (cased-based payment).
Fee-for-service. Arrangements for fee-for-service payment call for all health services to be priced separately and paid for separately. Under a health insurance reimbursement scheme based on fee-for-service, the more service that is performed, the higher the total reimbursement.
As a result, fee-for-service providers have an incentive to overproduce. If the insurer has purchasing power vis-à-vis providers, then fees can be negotiated downward or the government can fix fees, thus restraining production and total expenses. Japan, for example, has effectively restrained the growth of health expenditures relative to other developed economies through a system of fixed and regulated fees (Ikegami 1991). This, however, will distort relative prices over time leading to inefficiencies in the mix and types of services provided.
Capitation. Under a capitation scheme, providers of health care are paid a fixed amount per person per year to provide services. The most common capitation scheme in the United
States, is that used by health maintenance organizations (HMOs) (Andreano and Helminiak
1987; Luft 1981). Commercial insurers often refer to insurance contracts, which are financed through capitated payments, as “risk-products” because the financial risk of large medical expenses is borne by the contracted provider. Thailand’s employment-based health insurance scheme, under the Social Security Act of 1990, purchases service bundles for beneficiaries with capitated payments (Mills et al. 2000). Capitation is seen as a method for controlling costs as it removes the incentives inherent in a fee-for-service scheme to provide large amounts of services and testing and encourage utilization. Whereas with fee-for-service, greater service provision
97 generates larger revenues, under a capitation scheme each procedure contributes to costs but adds nothing to revenues because payments are made in a lump sum. Providers, therefore, have an incentive to be cost-efficient in the production of services and to restrain overuse. Theoretically, the most cost-efficient mixture of inputs and services will result under capitation. But providers have incentives to reduce costs by reducing their levels of service. In addition, providers also have incentives to encourage positive selectivity by attracting healthy patients (cream-skimming, cherry-picking) and discourage negative selectivity by avoiding ill patients (dumping)
(Newhouse 1996). These are major reasons why an effective health policy requires monitoring of contracted providers.
Prospective payment systems. Under a prospective payment system, fixed payments per episode or condition are determined before a medical condition occurs. This often referred to as case-based payment. For example, in the U.S. hospitals are paid on a per case basis under the
DRG prices (Fetter 1991). As with capitation, the lump-sum nature of the payment means that providers suffer economic losses or earn profits if the cost of service varies from the payment level. Also as with capitation, providers have an incentive to be cost-efficient and to restrain the use of services. A disadvantage is that hospitals also have an incentive to discharge patients too early. A major difficulty with this payment scheme is selecting the right payment levels.
Equity
Health insurance provides a vehicle for achieving equity goals. Most public insurance schemes explicitly incorporate cross-subsidization from high-income groups to low-income groups and medically indigent individuals as a feature of the plan. To ensure that all have needed coverage and access, subsidies must be provided to targeted beneficiary groups. These subsidies
98 are of two types. First, low-income groups need assistance in meeting their premium obligations.
Generally, public schemes incorporate income tests for premium or tax contributions. Second, some households have financial difficulty with co-payments and require additional subsidies at the point of service.
Equity issues have been carefully investigated for 12 OECD countries (Wagstaff and van
Doorslaer 1992; Wagstaff, van Doorslaer, van der Burg, et al. 1999; van Doorslaer and Wagstaff.
1992; van Doorslaer, Wagstaff, van der Burg, et al. 1999). The most recent evidence suggests that tax financing of the health care system is a progressive method of funding care (Wagstaff, van Doorslaer, van der Burg, et al. 1999). Social insurance is also progressive in the OECD, except for Germany and the Netherlands, where heavy reliance on insurance premiums, instead of taxes, has generated a regressive, albeit well financed, system. Direct (out-of-pocket) payment is the most regressive method of financing health care. These results suggest that the Korean system, where out-of-pocket payments are roughly 50% of financing, would tend to be regressive. By contrast, the Japan system, which is heavily reliant on general taxes, is likely to progressive. However, this issue has not been carefully and extensively researched for the ANE countries and therefore, definite conclusions cannot be drawn.
Group size
For financing schemes to be efficient and sustainable, critical mass must be achieved in the size of risk pools. Diamond (1992) suggests that groups of 10,000 to 100,000 individuals will be sufficiently large to make a generous package of preventive, promotive, and curative care financially viable. This group size may not be necessary to achieve more modest goals of financing simple prevention, basic drug therapy, and primary care, although under usual
99 circumstances, the larger the group, the easier it is to spread risk and predict expenditures. In the case of community financing, the public sector can act as a facilitator in pooling groups to achieve optimal size. This is particularly important in the case of community financing among rural towns and villages.
Mandated coverage
Financial sustainability of any insurance scheme requires that the design safeguard against adverse selection. The problem is most apparent when households with sick members and high expected expenses choose to participate in a finance scheme and healthy low-expense households choose not to participate. Voluntary systems inevitably face this problem. If a scheme is able to attain universal participation, the adverse selection problem is eliminated. This is why compulsory programs are frequently recommended by analysts.
Some analysts claim, however, that a voluntary health-insurance scheme at the village level in Africa has been successful in financing basic services (Chabot, Boal, and da Silva 1991).
In this case, social persuasion at the village level and government subsidy at the national level achieved high participation. Nevertheless, in the long run voluntary systems are fragile because rising expenses will cause healthy participants to exit or force an increase in government subsidies.
Catastrophic/disability insurance
The greatest underlying demand for health insurance is for protection against large, catastrophic expenses that are unpredictable (e.g., long-term disability). Unfortunately, the same characteristics that create the need, namely large expenses and lack of predictability, also make
100 catastrophic insurance the most difficult to underwrite and the least attractive for the private sector to supply. Large risks are best handled by large diversified groups, and the public sector may have certain advantages in facilitating large-group formation. If community financing schemes are expected to deliver acute care and be financially viable, the public sector must backstop catastrophic care through finance or referral.
Cost containment
Efficient financing requires two types of controls on costs. First, some level of cost- sharing by consumers of health services is efficient. Second, price controls on markets receiving the most generous financing are potentially efficient. Such pricing policies have successfully restrained expenditures in Japan (Ikegami 1991) and represent an important policy instrument for
Developing countries within the larger health financing scheme.
Employment-based insurance programs
Employment-based health insurance should be neither encouraged nor discouraged as a matter of public policy but rather should voluntarily arise by virtue of scale economies and similar sources of fundamental economic efficiency. South Korea, Japan, Taiwan and, to a lesser extent, the Philippines began their health insurance schemes by mandating coverage at the level of the employment relation initially with civil servants in all cases, and later with all workers in the case of South Korea, Japan and Taiwan. As a practical matter this may be the easiest way to operationalize either a public or a private insurance scheme. The long-run consequences of employment-based schemes include some undesirable outcomes, however, as illustrated in the case of the United States.
101 In the United States, most private health insurance is employment-based. Most large employers provide health insurance to employees as part of the total compensation package in the form of a fringe benefit. There are several economic reasons why employers and employees find that such an arrangement is to their mutual private advantage. First, relatively large groups can achieve economies of scale, to the extent that such economies exist, in organizing and administering health plans. In this regard the employment-based groups may also experience economies of scope in administration, as employers maintain administrative records on employees in the normal course of business. Thus, there may be cost complementarities in processing and maintaining information on employees and their families. Second, larger groups may have greater purchasing power than individuals in negotiating health insurance contracts and thus better able to obtain lower prices or better coverage. Third and most significant, employment-based health insurance coverage is treated as nontaxable income under U.S. tax law. As a result, employer-provided health insurance in afforded a substantial tax subsidy, which increases the demand for health insurance and in turn the demand for medical care (Feldstein and
Friedman 1977; Vogel 1980; Greenspan and Vogel 1980). This tax subsidy distorts both the quantity and the type of health insurance purchased (Pauly 1986), as well as labor market decisions and is thus a source of allocative inefficiency. Moreover, the tax subsidy afforded to employer-provided insurance is regressive in that it confers greater benefit to higher-income households, which face higher marginal tax rates. In sum, the U.S. employment-based system has distorted health care and labor markets and contributed to escalating costs, inefficiency, and inequity in the health care system.
In contrast to the U.S. system, the private finance system in Australia is based at the individual and household level and has no direct linkage to employment (Altman and Jackson
102 1991). Health insurance premiums, when paid by employers, are treated as taxable income. As a result, there is relatively little distortion in health insurance choice and labor market decisions.
The absence of an artificially imposed linkage between insurance and employment partly explains the lower expenditures of the Australian health care system relative to that in the United
States, 8.4% and 13.9% of GDP, respectively, in 1997 (OECD, 1999).
Analysts have recommended that the United States end the special tax treatment of health insurance premiums in a move toward efficiency (Diamond 1992; Pauly et al. 1991).
Nevertheless, the political economy of the health system makes it difficult to remove these employment-based subsidies, once granted (Wolfe 1993). This is why it is important for policymakers within developing countries to avoid creating strong employment-health finance linkages in their health systems or allowing those linkages to become entrenched.
Competition and regulation
For reasons already discussed, neither a purely privatized health system relying on unfettered competition nor a purely public system is likely to generate optimal efficiency and equity given a society's goals. The best of all feasible solutions is probably second best overall.
One potential solution is a system in which all households participate and make informed, but limited, choices among competing health insurers and health care providers. The government acts as a facilitator and rule- maker to limit the inefficiencies of competition. The government sector also subsidizes some households to achieve equity and acts as the insurer/financier-of-last- resort to secure the system. Such a system would fall under the general rubric of managed competition (Enthoven 1988) and has been proposed as a reform policy for the United States
(Diamond 1992; Pauly et al. 1991).
103 Labor and Retirement Policy
As populations age and labor force growth slows in Asia the labor force choices made by older workers and the employment practices they face will take on increased salience. Economic performance will be influenced in two ways. First, the burden of the old-age support system will be influenced by the retirement decisions made by older workers and their employers. If workers respond to increased life expectancy by extending their years in the work place, the material needs of the elderly can be met at a smaller cost to younger generations. If, on the other hand, workers retire at a younger age, additional resources will have to be mobilized for their support during retirement. Alternatively, living standards for the elderly will decline relative to those enjoyed by younger generations.
Second, the labor force behavior of older workers will become increasingly important as labor force growth slows and labor shortages emerge. In part, older workers will be more important simply because a larger share of the population and the workforce will be fifty or older. In addition, any response in the overall supply of workers will increasingly depend on decisions by older workers. Prime age adult males are for the most part fully employed. The high and increasing returns to education will provide a powerful incentive for young adults to opt for school and delay their entry to the workforce. Most Asian countries have been reluctant to open their borders to substantial numbers of foreign workers. Only increased labor force participation among women and older workers will moderate the influence of slower or negative growth of the working age population.
The employment decisions made by older workers will bear not only on aggregate economic performance, but also on their individual welfare and that of their families. For many
104 older workers continued employment is critical to maintaining economic security. For some, however, deteriorating health may dictate withdrawal from the labor force. Others may prefer an extended period of retirement at a reduced material standard of living.
To a great extent, retirement outcomes are the result of the interaction between labor demand and labor supply. Older workers make decisions about their continued employment based on their wages and non-pecuniary benefits, their wealth and non-labor income, their expectations about public and private economic support, and their assessment of current and future materials needs. Firms make decisions about employing older workers based on their current productivity, their future value, and the wages that must be paid to retain their services.
If the decisions of older workers and firms are not influenced by tax and retirement polices and if labor markets are well-functioning, there is little rationale for government intervention. The reality, however, is that tax and retirement policies have considerable influence and that labor markets are characterized by numerous rigidities that affect labor force outcomes.
There are many ways in which government policy can be improved to eliminate barriers to the continued employment of older workers and to mitigate some of the economic problems associated with rapid aging.
Retirement trends
The trend towards earlier retirement appears to be a persistent feature of development.
The experience is most thoroughly documented for the developed countries. Since 1950 the average retirement age for men has dropped in every OECD country but Iceland. Other countries have experienced a marked decline. In Finland, Spain, and the Netherlands the retirement age
105 has declined by more than 6 years. In Austria, Belgium, Finland, France, Luxembourg and the
Netherlands the average retirement age for men had dropped below 60 by 1995 (Table 24).
Table 24. Average retirement age among men, OECD countries 1950-1995
Comparable data are not available for most Asian countries and, consequently, the trends in retirement and the factors influencing work have been investigated less thoroughly (Hermalin,
1997). Nonetheless, trends in labor force participation rates suggest changes in Asia that are similar to those occurring in the West. Retirement trends in Asia are summarized below in two ways using labor force participation rates (activity rates) for men. The first is the median age of workers who retired during any five-year period. This measure is crude because it is based on broad age groups and does not capture non-linearities that characterize labor force participation.
It also reflects the age-distribution of the underlying population. Moreover, the upper age group is 65 and older. Frequently, more than half of this group is still active. In this case, the median age is extrapolated assuming a linear decline in activity rates.
The second methods uses the retirement hazard rate, the proportion retiring between age a-5 and age a out of those who were active at age a-5 (Hurd, 1990). Current retirement behavior is captured in crude fashion assuming that the cross-section activity rates capture the participation probabilities for individuals. The hazard rate from one age group, say 55-59, to the next age group is calculated as:
LFPR60-64 H55-59 to 60-64 =1 - LFPR55-59 where H represents the retirement hazard rate and LFPR represents labor force participation rates.
106 The median ages at retirement, presented in Table 25, exhibit two interesting features.
First, most Asian countries are experiencing a substantial decline in the median age at retirement.
Although the concept of the median age at retirement is different from that of the average retirement age, the changes in Asia appear to be similar in magnitude to the changes in OECD countries. Second, the cross-sectional pattern is broadly consistent with the view that development leads to an earlier age at retirement. The median age at retirement for East Asia had dropped below 65 by 1980 and reached 62 for 1995. In contrast, the median age was close to 65 in both South and Southeast Asia in 1995. Retirement may not be an option in a poor rural setting—many people work as long as they are able.
Table 25. Median age at retirement, Asian males aged 40 and over 1950-2010
Hazard rates for Asian male for 1950-2010, shown in Figure 25, confirm the shift towards an earlier age at retirement. The probability of retiring between the ages of 55-59 and
60-64 given membership in the labor force at age 55-59 was 0.251 for Asian men in 1950. In other words, about one quarter of men aged 55-59 in 1950 retired by the time they reached 60-
64. The retirement hazard rate has increased rapidly for the last 50 years and is projected to increase for the next few decades. The probability of retiring between 55-59 and 60-64 increased to 0.373 in 1990, and is projected to reach slightly less than 0.5 by 2010.
Figure 25. Retirement hazard rates, Asian men 1950-2010
Causes of retirement
107 The clear connection between the level of development and age at retirement suggests a simple model of retirement behavior, namely that workers are opting for a longer retirement because they can afford it. Most research on retirement behavior provides some limited support of this view, but the research also points to a variety of forces that are influencing behavior. Most recently, Auer and Fortuny (2000) used the 1995 European Union Labour Force Survey, to examine why retired men aged 55-64 left their last job (Table 26). Although there is a wide variation among countries, there are two noticeable causes of early retirement. First, a substantial percentage of men reported that their retirement was voluntary. Almost one half of the men in
Austria and 30 percent of the men in Denmark offered this explanation. Second, in countries experiencing deep recessions in 1990s, namely Sweden, Finland, UK and Spain, high percentages of older men were dismissed from their last job. These results and many other studies demonstrate that a comprehensive understanding of retirement must look to three set of factors: supply factors that influence the willingness of older workers to remain in the labor force; demand factors that influence the willingness of employers to hire or to retain older workers; and, institutional rigidities in the labor market that preclude outcomes that would be preferred both by older workers and their employers.
Table 26. Main reasons for leaving the job, men aged 55-64, EU
Labor supply effects
The rise in wages that accompanies development has conflicting effects on retirement behavior. Retirement becomes more expensive to workers in the sense that they forego a higher wage by withdrawing from the labor force. However, higher wages over the work-life provide
108 individuals with the resources needed to support a longer retirement period. The trend in retirement age reflects, in part, the strength of these two effects in the face of general wage growth, but also the impact of public policy on wages and retirement wealth.
First, governments impose taxes on income and/or earnings reducing the after-tax wage and, consequently, the cost of early retirement. In some instances, the effective tax rate faced by those near retirement age may be higher than the tax rate faced by other workers because the lifetime value of retirement benefits are lower to those who continue to work. To the extent, then, that these policies exist and encourage early retirement, government policy serves to undermine economic growth and support systems for the elderly. Second, public transfer programs shift wealth across generations. In principle, transfer programs can favor any particular generation, but in practice public programs have transferred wealth from those who were young workers or who had not yet entered the labor force, including those not yet born, at the time the program was established to older workers and retirees. During the early stages of transfer programs, then, the wealth effect reinforces the effect of reduced wages encouraging a younger age at retirement. When transfer programs mature, however, their impact on retirement is ambiguous since older workers face lower wages and lower wealth than they would have in the absence of the transfer program. What is unknown, however, is whether or not the effects on retirement behavior are negligible, substantial, or some where in between.
In developed countries social security benefits and private pensions have increased over time and are now the primary income sources for the elderly. In the US, for example, the income of the elderly has grown more rapidly than the income of the non-elderly or the income of households to which children belong. This has been a direct outcome of increases in the Social
Security program. Income from earnings fell from 29 percent in 1967 to 17 percent in 1986. The
109 fraction of income from social security and pensions increased from 49 to 54 percent, mostly because of increases in Social Security. Several researchers have concluded that large increase in
Social Security benefits were responsible, in part, for fall in labor force participation rate among older Americans (Hurd, 1990).
Many other studies have investigated the effect of pension and social security programs on retirement decision. Most conclude that the growth of pension and social security programs has led to earlier retirement, but no consensus has emerged about the magnitude of the effect. For example, Gustman et al. (1994) find small effects; Lumsdaine et al. (1997) substantial ones.
Studies that have focused on social security rule modifications conclude that their impact is modest. Simulations using firm data show that changes in social security policy in the U.S. will have little impact on the age at which workers accept their private pension, mainly because private pension wealth in the U.S. is substantially larger than social security benefits (Stock and
Wise, 1990; Lumsdaine et. al., 1997). Of course, many workers in the U.S. do not have private pensions on which they can draw. Studies of the impact of pensions on retirement usually find that workers with generous pensions retire earlier than those with smaller pensions, but the magnitudes of the effects vary considerably across studies.
Evidence for Asian countries is very limited. Asian firms have used compensation schemes to encourage older workers to retire early. In Korea, for example, as unemployment became more acute during the recent financial crisis a variety of programs were introduced to encourage early retirement. A number of firms offered compensation for loss of earnings and to bridge the gap until a full pension could be claimed. Although there is little empirical evidence on the magnitude of the impact, reliance on the programs suggests that firms perceive their employees as being responsive to such incentives. If so, these measures might have accentuated
110 the decline in the retirement age. Expansion of public programs currently being considered in many countries may lead to further declines in the age at retirement.
Labor demand and retirement
Retirement decisions are also affected by labor demand. Many Asian economies, especially in East and Southeast Asia, experienced strong economic performance, until recently, and a correspondingly strong demand for workers. In recent years, however, economic crisis and slower growth in many Asian countries has adversely affected the demand for labor. Women and older workers are most vulnerable when enterprises restructure and few governments have effective policies that prohibit employment practices that discriminate on the basis of gender or age.
Many governments allow firms to target the elderly in the belief that it will improve job prospects for young employed males viewed as the primary breadwinner for the family.
Moreover, because of the seniority-based wage system in many Asian countries, employers may believe that older worker are receiving wages and benefits that are too high relative to their productivity. Although older workers are more experienced than younger workers, they may have poor health and suffer other effects of aging that reduces their productivity. Consequently, the elderly may be forced into early retirement when there are general down-turns in the economy or when particular sectors or firms decline or restructure their production processes.
Estimating labor demand models has proven to be much more difficult than estimating labor supply models in part because of the lack of firm-level data. Consequently, evidence about the impact of demand side factors is relatively poor (Lumsdaine and Mitchell, 1999). One issue examined in empirical studies is the relative productivity of younger and older workers.
111 Although many economists, especially young economists, have assumed that younger worker are more productive, other studies find that older workers have superior skills (Mitchell, 1990).
Although older workers may be less productive than younger peers due to skill obsolescence, this is not a consequence of aging per se but due to a lack of re-training (Auer and Fortuny,
2000).
Most studies find that poor health leads to early retirement (Rust, 1988; Quinn et. al.,
1990). Poor health raises the opportunity cost of work, influencing the supply of labor. It also signals lower productivity to employers affecting the demand. Given the improvements in health that have accompanied economic development, it is puzzling that age at retirement has declined, not increased. It may be that other factors have overwhelmed the influence of better health. Or, it may be that higher incomes have enabled workers with poorer health to withdraw from the labor force.
It is far from clear that governments should support practices aimed at encouraging early retirement by workers. A reduction in the employment of older workers does not necessarily lead to an increase in the employment of young workers in the short run, because old and young workers frequently have different skills. Moreover, entry to and exit from the labor market do not occur in the same sectors or occupations. In the long run, withdrawal of older workers from the labor force does not necessarily result in greater employment of younger workers either. The number of jobs is not fixed in the long run. From a social welfare perspective, encouraging early retirement may reduce employment, income, and economic welfare, in general.
112 Rigidities in labor market
Labor market rigidities refer to structural impediments created by firms or labor unions or imposed by governments that reduce the flexibility of employment practices. The presence of these rigidities may lead to welfare losses by reducing employment by older workers. The rigidities can take a variety of forms. Many governments in Asia and throughout the world have mandatory retirement ages. Firms may have inflexible rules about work hours. Seniority wage systems may preclude the downward adjustment of salaries for workers with diminished productivity.
To a great extent, labor market rigidities are a feature of the growth of the urban, industrial sector. Agricultural workers, the self-employed, and those working in small-scale enterprise are much less subject to market rigidities than are those working in the public sector or for large-scale industrial firms. Thus, the shift of labor from the rural agriculture sector to the urban non-agricultural sector has led to increased labor market rigidities and possibly a younger age at retirement. Although other factors, emphasized above, have played a role in the decline in the age at retirement, the simple relationship between activity rates of elderly and the percentage of agriculture sector employment reinforces the emergence of labor market rigidities (Figure 26).
Figure 26. Relationship between activity rates of elderly and the percentage of employment in agriculture sector, 1990
Older workers approaching retirement often desire gradual retirement because the desired amount of working time usually decreases gradually as tastes shift toward leisure. If the labor market is flexible and older workers can vary their amount of work according to their preference
113 and productivity, then either a fall in their desire to work or a fall in their marginal productivity will result in a fall in the volume of work. However, what has been actually happening is somewhat different because of labor market rigidities. The evidence for the U.S. is illustrative.
About 70 percent of the transitions in the U.S. have been from full-time work to no work (Rust,
1989; Berkovec and Stern, 1988). Several studies have found marked peaks in retirement ages rather than a smooth transition (Rust, 1989; Lumsdaine et. al., 1996). Researchers suggest that it is “institutional rigidities” in labor market that is responsible for this particular pattern of retirement (Lumsdaine and Mitchell, 1999). These rigidities include the existence of statutory retirement age (pension-receivable age) and working time constraint (Lumsdaine and Mitchell,
1999).
Mandatory retirement ages are pervasive. Most countries in the world impose a statutory age at retirement (Table 27). Some countries have a lower retirement age for women than for men. This is a puzzling form of gender discrimination given that women have a longer life expectancy than men in most countries of the world. Many countries have moved to equalize the retirement age for men and women. The statutory retirement age in less developed regions is generally lower than in more developed regions. Developed countries usually have a statutory age of either 65 or more. For most Asian countries, the statutory retirement age is less than 65. It is 65 in Japan, 60 in the Philippines and Korea, and 55 in Indonesia and India. The lower mandatory retirement age in developing countries may reflect the lower life expectancy and poorer health status of older workers, but many developing countries have been slow to adjust mandatory retirement ages upward despite rapid improvements in life expectancy and health status.
114 Table 27. Statutory retirement age
The impact of mandatory retirement laws on work by the elderly depends, of course, on the extent to which the policies conflict with the desires of older workers and their employers.
Several years ago the U.S. passed a law banning discrimination on the basis of age. The elimination of a mandatory retirement age had a strong, positive impact on employment rates for older workers. Moreover, the rise of older workers’ employment did not result in a decline in employment among younger workers (Neumark and Stock, 1999).
Mandatory retirement is the most obvious example of a labor market rigidity that influences retirement decisions, but other features of the labor market may also play an important role. The market rigidities sometimes include the inability to change working hours on a given job or changing to a different job that offers the desired combination of hours (Hurd, 1996). In a rigid market, the optimal constrained choice can be to work more than is desired and then not to work at all, rather than to reduce work effort more gradually. Firms that have seniority based wage profiles may encourage retirement rather than reduce either wages or the amount of work when wages exceed older workers’ marginal productivity (Lazear, 1979). Some Asian countries, especially Japan and Korea, have had a long tradition of seniority based wages, that might encourage earlier retirement.
Concluding remarks: policy options
Policies that undermine work effort and promote, or dictate, early retirement are damaging in several ways. First, older workers who are not yet financially prepared for retirement are forced to accept a lower material standard of living during their retirement years.
115 Given the importance of the family support system in Asia, some of the extra burden will be shouldered by the families of the elderly. Second, economic growth is reduced by the loss of human capital. Third, the fiscal viability of public support programs is damaged by the decline in the number of earners and the tax base relative to the number of beneficiaries. These issues are particularly salient in aging societies, but under any circumstance eliminating work dis- incentives and labor market impediments is sound economic policy.
Different policy options address the impact of pension programs, tax policy, and labor market rigidities. One option is to increase the mandatory retirement age or, preferably, to eliminate it altogether. For many workers the pensionable age, the age at which an individual qualifies for a pension, affects retirement behavior with as much force as the mandatory retirement age. One possible approach is to remove the statutory pensionable age and to provide pension benefits that are actuarially determined by the age at which an individual chooses to retire. Thus, a decision to retire early does not impose additional costs on the retirement system nor does the retirement system either encourage or discourage retirement by older workers.
Although income and earnings taxes influence work incentives and, in many instances, reduce work effort, reducing average tax rates may not be a viable option. However, adopting a
PAYGO pension program necessarily requires higher taxes with adverse effects on work effort.
Another policy option is changing the tax structure. One possibility is to shift the relative burden born by labor and non-labor income. Of course, increasing taxes on non-labor income will adversely affect saving and investment. Another option is to flatten the rate structure. In a progressive system, the marginal tax rate rises as workers increase the hours employed. Hence, the after-tax wage declines the more individuals work. Flattening the tax structure will generally encourage employees to work more hours. In the face of labor market rigidities that limit the
116 availability of part-time employment, a flattening of the tax rate may lead workers to increase their age at retirement.
The elimination of labor market rigidities is a high priority for effectively utilizing the human resource potential of older workers and for allowing older workers to remain in the labor force as long as they so choose. If employers have the flexibility to hire older workers on a part- time basis, vary responsibilities as the capabilities of older workers change, and pay a wage commensurate with the productivity, not the seniority, of older workers, employers will be much less reluctant to hire or retain older workers. Gradual retirement has some clear advantages. It enables older workers to postpone the complete loss of labor earnings. It allows employers and employees to adjust the demands of work to any decline in capabilities that might be associated with age. It also allows firms to retain experienced workers and facilitates the transmission of that experience to the next generation of employees. Particularly as labor force growth slows and, in some countries, turns negative, more flexible and innovative employment practices will become increasingly attractive.
In addition to eliminating rigidities in the labor market, more active steps are possible to meet the human resource needs of older workers. Occupational re-training programs and general educational upgrading will allow older workers to take up new occupations and to cope with rapid technological change. It is well known that individuals with low educational attainment and low skills are at a higher risk of remaining jobless for an extended period of time. The education upgrading and re-training will provide the basis for workers to acquire skills throughout their working life and thus enter their older ages relatively well equipped. Without re-training, skills of the elderly will be obsolete and the incoming cohorts of younger workers will continue to have skill advantages. Employers may be reluctant to invest in the human capital of older
117 workers because of concerns that the employer will derive benefits over a shorter period of time.
Turnover among older workers, however, may be much lower than it is among younger workers.
Improving the flexibility of the labor market, of course, raises dangers that older workers will experience age discrimination with their levels of responsibility and/or wages reduced for reasons unrelated to their capability. The need for effective systems for monitoring and correcting age-discrimination will become increasingly important if systematic abuse is to be avoided.
Old-age Support Systems
Family support systems
In most traditional Asian societies, the family bears primary or exclusive responsibility for providing old age security. Most elderly live in an extended, multi-generation family and rely on their adult children, their spouse, and other family members for both material needs and personal care. Even when children are living separately from their parents, contact is more frequent and resource flows more substantial than in the West. By and large, the family-based approach has been remarkably successful, especially given the potential for great inter- generational income inequality in the rapidly growing economies of Asia. When the elderly were young, their earnings were a pittance in comparison to what today's workers can command. But the family support system has from every indication insured that the elderly have shared in the region's economic success. Income inequality is quite low among developing countries, particularly in East and South Asia, and the elderly appear to have maintained standards of living comparable to those enjoyed by their offspring.
Despite its success, the traditional family support system is under a great deal of pressure from demographic, social, and economic change and these pressures are likely to grow in the
118 coming decades. The same demographic forces that are leading to a decline in the aggregate support ratio are leading to a decline in the support ratio at the family level. With increases in life expectancy adults are more likely to have surviving elderly parents, and the persistence of low fertility means that sons and daughters will have fewer siblings with whom to share the responsibilities of parental care. Economic forces are reinforcing the demographic ones. The young are leaving their home for urban centers and new employment opportunities leading to more physical isolation from their parents. Women are entering the workforce, reducing the time available to provide family support. Exposure to the West may be introducing new ideas about marriage, family, and individualism. Marriage rates have dropped substantially in a number of countries. If a significant number of men and women choose to forego marriage altogether, this could have significant implications for family support. Finally, the development of financial markets and employee-based and public pension programs offer alternative mechanisms for securing old-age support.
The Asian family is responding to these pressures. Change has been most pronounced in
Japan and recently in South Korea. In other countries the changes have been relatively modest to this point. Throughout the region, however, the family support system remains much stronger than in the West. The issue for public policy is not to devise a support system that presumes a minimal role for the family. Rather the challenge is to assess the distinctive advantages of alternative approaches and to devise complementary programs.
The current situation
In Asia, most elderly live with their children. Census and survey data available for
Thailand, the Philippines, Indonesia, Sri Lanka, Singapore, Taiwan, Vietnam, Korea, and Japan
119 confirm that the extended, multi-generation family is a resilient feature of Asian society. Japan and South Korea are distinctive among these countries in that only half of their elderly were living with children in 1990. In the other countries, the percentage of elderly living with children ranged from 67 percent in Indonesia to 85 percent in Singapore (Table 28). Intergenerational co- residence is much less frequent in the West. In Australia, the Netherlands, and the United States only 16, 12, and 13 percent, respectively of those 65 and older lived with their children in the
1980s (World Bank, 1994).
Table 28. Living arrangements, elderly population, selected Asian countries.
The elderly do not depend exclusively on their children. Many living independently of their children are living with their spouse. Older men are more likely to be living with their spouse as compared with older women in every country for which data are available. The difference is particularly pronounced in Japan where 36 percent of elderly men live only with their spouse as compared with 16 percent of elderly women. The differences between men and women reflect the reality that most women outlive their husbands both because they are younger than their husbands and because they live longer. Because older women are much more likely to be widowed than older men, they are also much more likely to live alone. In Korea, for example,
13 percent of women 65 and older lived by themselves as compared with only 3 percent of elderly men. In Indonesia, 15 percent of elderly women lived alone and only 3 percent of elderly men. In Thailand 5.5 percent of elderly women lived alone, as compared with 3.3 percent of elderly men (Table 29). Very small percentages of either older women or older men live in
120 households consisting of unrelated members. Relatively small numbers live in institutions. Only in Japan are elderly women more likely to be living in institutions than are elderly men.
Table 29. Non-Family Living Arrangements for the Elderly (65 and older) in Asia.
The elderly are far from a homogeneous group. Men and women differ in their needs and the way that they rely on their family. Likewise, the situation for those in their sixties is considerably different than for those in their seventies and eighties. Detailed data are more limited at older ages because of the small samples in many surveys. However, survey data from the 1980s for the Philippines, Singapore, Taiwan, and Thailand show that those who were 75 and older were actually somewhat less likely to be living with their children than those in their sixties. Interpreting these data is somewhat problematic. The decline in co-residence with children may reflect nothing more than the fact that children are becoming adults and no longer depending on their parents. Despite the decline a high percentage of those 75 and older are living with their children (Table 30).
Table 30. Percentage living with one or more children by parents' age, ASEAN survey.
A somewhat different picture emerges from data on the proportion living alone drawn from censuses (Table 31). Reliance on census data allows greater age detail and provides clear evidence, especially for women, that the proportion living alone reaches a peak and then declines at older ages. For men, the proportion living alone typically peaks among those aged 80-84, although a less uniform picture emerges in the Philippines and Thailand. For females, the
121 residents. The urban survey was drawn from locations near Islamabad and Rawalpindi; the rural sample from villages in Punjab and N.W.F.P. Hence, it is not representative of Pakistan as a whole. Of rural residents, 86 percent of those 60-69 and 93 percent of those 70 and older lived with an adult child (18 or older). Of urban elderly, either 60-69 or 70 and older, 92 percent lived with and adult child (Afzal, 2000).
National level data on household size provides a rough indication of living arrangements in South Asia and these data are consistent with the view that the extended family persists in
South Asia. The number of adults 15 and older per household in South Asian countries ranges from 3.0 members in Bangladesh to 3.8 members in Pakistan. In most Southeast Asian countries, the number of adults per household are comparable. In Malaysia, Myanmar, the Philippines,
Singapore, and Thailand the average household has more than 3 adults. Among East Asian countries the number of adults per household is under 3 adults per household ranging down to
2.5 in Japan in 1995 (Table 32). One cannot directly infer living arrangements for the elderly from such a crude measure as adults per household. The patterns we observe in the region partly reflect differences in age structure and differences in living arrangements among young and middle-aged adults. However, the data suggest that the extended family is as prevalent in South
Asia as it is in Southeast Asia.
Table 32. Household size, children and adults per household, Asian countries.
Despite the continuing prevalence of the multi-generation extended family, there are clear indications of a shift toward more independent living arrangements. The change has proceeded much further in Japan and Korea than elsewhere. In 1950, 80 percent of elderly Japanese lived
123 with their children (Ogawa and Retherford, 1997) as compared with 50 percent in 1990. In South
Korea, the proportion of elderly women living alone nearly doubled between 1980 and 1990 and the proportion living with their children dropped from 78 percent in 1984 to 47 percent in 1994.
In Taiwan, between 1973 and 1986, the proportion of older parents living with a married son declined from 82 percent to 70 percent (Weinstein and others 1994). Outside of East Asia, living arrangements are changing more slowly or not at all. In Thailand and Singapore, the percentages of older adults living with a child have dropped modestly. No change is evident in the
Philippines. The number of adults per household does not appear to have dropped at all in
Southeast or South Asia (Table 32).
The changes in living arrangements that are occurring can only be taken as indicative of changes in the family support system. One difficulty is that resource flows within the household are rarely documented; hence, one cannot with any degree of assurance assess whether co- residence is a means by which children are providing support to elderly parents, by which elderly parents are providing resources to their adult children, or neither. Thus, declining coresidence does not necessarily indicate a decline in support for the elderly.
A few studies in Asia have tried to assess the extent to which co-residence was accompanied by resource flows. In a 1988 survey, adult children in Japan were found to be much more much likely to move into their parents domicile than the converse, suggesting a transfer of housing services from parent to child. However, the reported circumstances surrounding decisions to co-reside indicated that children moved in with their parents in response to some perceived need on the part of the parents (Martin and Tsuya, 1994.) In a 1986 survey, children in
Thailand were asked about the support that they provided to their elderly parents. Among
124 coresident children, 32% provided money and 55% food and clothing and among non-coresident children, 41% provided money and 30% food and clothing (Knodel and Chayovan, 1997).
A second difficulty is that the extent to which living arrangements are changing may be exaggerated by the simple dichotomy "living together/not living together" implicit in the definition of a household. In Thailand, for example, members of the same extended family may live in separate households, but within a single compound. In an urban variant on the same theme, members of extended families in Singapore live in separate units within the same apartment building (Yap, 1998). In Tokyo, 50.9% of those 65 and older lived with their children in 1989, but another 11% lived in the same complex or the same domicile but maintained a separate budget. An additional, 11.1% lived in the same neighborhood as one of their children
(Miyajima, 1994).
Despite these complexities there are clear indications that the family support system is declining or expected to decline in the future. Ministry of Labor data for Japan reports the proportion of elderly dependent "primarily on their children's income" in 1980, 1983, and 1988.
Thirteen to fourteen percent of working men in each year depended on their children for the primary source of income. Among retired men, the proportion depending on children as the primary source of income declined from 37% in 1980 to 23% in 1988. Among working women, the percentage dependent on children declined from 37% in 1980 to 28% in 1988. And for women who were not working, the percentage depending on their children dropped from 52% to
24% in only eight years (Miyajima, 1994).
Surveys of young Japanese adults indicate that they are increasingly likely to discount the family as an important support system on which they can rely in their old age. Remarkable documentation about shifting attitudes is provided by a survey of ever-married women under the
125 age of 50 conducted by the Mainichi Newspaper. Since 1950, women have periodically been asked: "Are you planning to depend on your children in your older age (including adopted children, if any)?" In 1950, 65% replied that they expected to depend on their children, 16% responded that they had never thought about the issue, and 20% reported that they did not expect to depend on their children. By 1990, only 18% responded that they expected to rely on their children, 20% had never thought about it, and 62% did not expect to depend on their children
(Ogawa and Retherford, 1993).
Taiwan's family support system has been subject to less erosion. The proportion of married women aged 20-39 reporting that they regularly gave money to the husband's parents increased from 32% in 1973, to 42% in 1980 where it remained in 1986. The percentage reporting that they never gave money to their husband's parents declined from 35% in 1973 to
15% in 1980 and 1986. The proportion reporting that they never gave money to the wife's parents declined from 58% in 1980 to 48% in 1986. (The question was not posed in
1973)(Weinstein et al., 1994). However, recent analysis of household income for Taiwan indicates that beginning in the late 1980s the per capita income of the households in which elderly live has declined relative to households containing no elderly. The source of the decline was an increase in the extent to which the elderly were living independently of their children.
Even though interhousehold transfers to the elderly have increased there, they have been insufficient to offset the decline due to changing co-residence (Mason and Miller, 1996).
Other Asian countries exhibit similar shifts. In South Korea, only 8 percent of women responding to a 1997 survey indicated that they wished to live with their children in old age and
70 percent did not (Lee, 1998). In the Philippines, recent survey results indicate that fewer adults wish to live with their children than is currently the case (Natividad and Cruz, 1997).
126 The future of family support systems
A variety of forces have been identified that are undermining and will continue to undermine the family support system. The decline in agriculture, urbanization, secularization, financial sector development, public policy, and demographics all seem to push in favor of formal systems of support for the elderly. In opinion surveys in several Asian countries described above respondents clearly expect to rely less on their family for old-age support than is now the case. In policy circles, eventual decline in the importance of the family appears to be widely accepted. The World Bank expressed this view succinctly: "It is inevitable that formal systems should eventually replace informal systems as the dominant form of old age support." (World
Bank 1994).
A review of the record over the last few decades does give one pause in reaching conclusions about how rapidly support systems will evolve in Asia and whether or not they will come to resemble those found in the West. The most economically advanced countries of Asia,
Japan, Singapore, Taiwan, and Korea, have already experienced many of the economic and social forces thought to undermine family support systems. Their populations are wealthy, educated, and urbanized. Their labor forces are employed in industrial and service sectors. They have well-developed financial sectors and, in some countries, well-developed pension programs.
Their populations are aging as rapidly as in the West. Yet, the family support system remains much more important in Asia. Perhaps, then, the family support system will continue to play a more important role than envisioned by many. The need for public programs is further diminished by the high rates of saving in many Asian.
127 That is not to say that the family support system will remain as important as it is today.
Many of the countries of Asia have yet to experience the economic and social changes that have occurred in the most advanced countries of Asia. Moreover, demographic changes on the horizon could prove to be very important. Among the countries of Asia, only Japan has experienced the
"demographic squeeze" caused by the combined impact of longer life expectancy and lower fertility. In the other countries, fertility decline has been quite recent and the elderly of today have historically high numbers of surviving offspring upon whom they can rely (Feeney and
Mason, 1998). Over the coming decades, however, elderly will be living longer and have fewer children, placing considerable strain on the family support system.
At the same time that demographics are working against support from children, they are working in favor of support by marital partners. This may be a particularly important change for elderly women who are so likely to be widowed. As the projections presented above show, the proportion widowed should decline substantially in the future increasing the extent to which the elderly can rely on their marital partners.
Public policy may also have an important influence on the role of the family in the future.
Several Asian governments have explicitly recognized the desirably of maintaining a strong family support system and have adopted pro-family policies. In Singapore, children are now legally responsible for the support of their elderly parents. In Malaysia and Singapore, public housing policy has been revised to better accommodate multi-generational living arrangements.
Many East and Southeast Asian countries are providing day care and other support services aimed at helping children care for their elderly parents. Malaysia provides tax incentives and other governments have considered or adopted similar policies (World Bank, 1994).
128 The advantages and disadvantages of family support systems
Enforceability: Will systems persist?
Except perhaps in Singapore the "contract" governing family support systems is a social contract not a legal one. As a consequence, norms and social pressure are relied on to insure that each generation cares for the preceding one. If some children choose not to honor the contract, their parents have no recourse and no support. If a generation chooses not to honor the contract, either a generation of elderly will suffer the consequences or the public sector will have to step into the gap.
This particular problem is not unique to a family support system. Businesses can fail leaving their employees pensionless. Governments can fall or renege on their commitments.
Hence, the relative advantage of formal and informal systems depends on the stability of the political and economic environment in which the system operates. In many settings, the family is the most reliable institution upon which the elderly can depend.
Risk pooling
Support systems for the elderly involve risk pooling at which family is disadvantaged.
The family consists of only a few earners who in many instances live in the same geographic locale and who are employed in the same occupation and industry. The elderly who depend on them will be subject to income shocks that can be averaged away in large-scale systems.
Similarly, in family support systems the members are exposed to the risk that their elderly members will live to a late age. Although in principle annuities can be relied on to insure against this risk, either these markets are under-developed or families choose not to protect themselves.
Either a large-scale public or private system can pool against this risk through its pension
129 program. It is common in Asia, however, to make lump-sum payments at the time of retirement.
By doing so, these programs do not insure participants against the risk of unusual longevity.
Administrative efficiency
Formal support systems are difficult to administer and subject to ineptitude and corruption. Thus, they may be beyond the capacity of countries lacking the appropriate human resources and legal and regulatory environment. In contrast, family support systems are extremely easy to administer.
Economic efficiency and growth
The tax and benefit structure of some formal systems can lead to economic inefficiency by discouraging work and saving. Formal systems are also susceptible to evasion and shirking.
To some extent these problems are less severe in a family support system because family members can more effectively monitor each other's behavior and because some costs are internalized.
Family support systems, however, also undermine the pension motive for saving and, consequently, may reduce rates of saving and growth (as compared, for example, with a compulsory, funded program such as found in Malaysia or Singapore).
Rates of return
Mature transfer systems, whether they are family based or public programs, can sustain rates of return no greater than the rate of growth of the economy. If workers accumulate wealth to finance their retirement, acting either independently or as part of a funded public pension
130 program, the rate of return is equal to the return to private capital. In many Asian countries, rates of economic growth reached unprecedented levels during the last few decades and the elderly enjoyed high rates of return through their participation in family support systems. In other Asian economies and the mature economies of the West, rates of economic growth have been much more modest. Consequently, the rates of return to transfer systems have been low as compared with those available through funded programs. The differences in the rate of return has been one of the key factors that has increased interest in the U.S. of moving away from its PAYGO Social
Security system.
Transition costs and unfunded liabilities
Family support systems are just like PAYGO public systems in another important respect. Both systems have unfunded liabilities. When few adults reach old age and even fewer retire, the liabilities implicit in a transfer system are relatively small. But as the retired population grows relative to the working age population, the debt of the support system becomes very large.
As countries in Latin America and the West shifting from PAYGO to funded pension programs these debts come due. Either they must be paid by current and future retirees who will forego some of the pensions that were committed to them. Or, if benefits are maintained, the more typical response, they must be funded by higher taxes on current and future workers.
If Asian countries maintain family support systems, the same logic applies. During any transition either benefits to those who are currently retired or who are about to retire will be sacrificed; or, current and future workers will bear the cost because they will be required to support their parents at the same time that they must save more to support their own retirement.
131 A recent analysis of Taiwan concluded that the unfunded liabilities required to maintain a family support system would reach five times national output by mid-century as compared with twice national output today. The rise in unfunded liabilities would greatly impede pension reform.
Public support systems
In the industrial economies of the West and the developing economies of Latin America, governments have been actively involved in insuring economic support for the elderly. In most
Asian countries, however, the public sector has played a relatively small part in the provision of old-age security. Asia has relied much more heavily on the traditional system of family support.
The trend, however, is toward more formal systems of social protection in response to or in anticipation of eroding family support systems. As the demographic transition proceeds and the family-level support ratio declines, providing old-age security will become an increasingly burdensome family responsibility. Other social and economic forces are reinforcing the demographic effects. Continued economic growth and Westernization are bringing about a shift in attitudes toward more individualistic lifestyles and consumption patterns. The development of capital markets is offering new investment opportunities allowing the elderly to rely on accumulated wealth rather than on their families for old age security. As noted in the previous section, the family support system has resisted these forces, but clear changes are underway in some countries. One of the major challenges facing the region is to respond to the changes in the family support system without accelerating its demise.
Government involvement in the provision of old-age support is justified primarily on two grounds. First, government programs address the growing life-cycle problem. Adults are typically living many years after they are no longer economically productive. As the importance
132 of family support declines, the elderly may increasingly rely on personal savings to finance consumption during their retirement years. However, many may not be sufficiently far-sighted to save adequately. Or, financial institutions may be inadequate to provide attractive and secure long-term investment opportunities. Second, governments may intervene out of concern over poverty and income inequality. In most Asian countries, income inequality is relatively low as compared to countries at similar development stages. However, as Asian countries have achieved higher average standards of living, many are re-examining their policies towards poverty and inequality. The recent financial crisis has fueled these concerns, although a clear assessment of the impact of the crisis on older populations is not yet available.
The actual policies that countries pursue will be influenced by both of these considerations. Indeed, they are complementary in the sense that policies that help working age adults prepare for retirement may help to alleviate future inequality. The evolution of policies towards the elderly may also be heavily influenced by political considerations, especially as the voting power and influence of the elderly population increases over the coming decades.
What policy choices are desirable for old age support systems in Asia? To address this question, we begin with a brief description of public support systems found around the world and a comparison of Asia’s programs to those found elsewhere. Then we address four important policy issues: (1) the extent to which governments should rely on mandatory programs; (2) alternative approaches to finance; (3) strategies for risk-sharing; and (4) private versus public management of programs.
133 Current situation
During the last sixty years national governments throughout the world have come to play an increasingly important role in providing for the social security of their citizens. Between 1940 and 1999, the number of countries offering social security programs increased from 57 to 172.
Old age, disability, and survivor benefit programs (OASDI) and work injury programs are the most common. Among those countries with any type of social security program, the number providing OASDI has increased from 33 (58%) to 167 (98%). Currently, only 5 out of the 172 countries with any type of social security programs do not offer old age benefit schemes. Asian governments have also adopted a wide range of social security programs. Only Bangladesh and
Myanmar do not offer OASDI in any form (U.S. Social Security Administration, 1999).
OASDI programs vary considerably with respect to the methods of finance, determination of benefits, and eligibility requirements. Most countries rely on contributory programs, i.e., programs that are partially or fully financed by payroll taxes. A number of countries in Europe and Asia rely on non-contributory schemes financed out of general revenues.
Most of these programs provide means-tested benefits; a few provide universal flat-rate benefits.
In many more countries retirement benefits are tied to the earnings history of each worker. In seven countries, national governments are not directly involved in providing benefits to the elderly, but mandate that private companies provide programs for their employees. Thirty countries, including Indonesia, India, Nepal, Malaysia, Singapore, and Sri Lanka in Asia, have mandatory saving programs, two-thirds of which are administered by the public sector.
Table 33. Types of schemes providing cash benefits to the aged, disabled and/or survivors, 1999
134 Although most countries in the world offer OASDI programs, in many countries coverage is limited to a small percentage of the work force. A few Asian countries, Malaysia,
Singapore, and Japan, have programs with close to universal coverage. Most Asian countries do not. The percentage of workers covered by old-age benefits in 1992 was only one percent in
India, seven percent in Indonesia, and 26 percent in Korea (Table 34). Many countries restrict coverage to those who are employed in the formal sector, who hold a regular job position, or who work for firms with a minimum number of employees or earnings. The Employees
Provident Fund in India, for example, applies a combination of provisions to restrict membership: employment must fall within the scope of one of 177 prescribed occupations; the establishment must have at least 20 workers; and, the establishment must have been operating for at least three years.
Table 34. Coverage of schemes providing cash benefits to the aged, disabled and/or survivals
The relatively modest scope of programs in most Asian countries is also reflected in financial measures of their size. The proportions of public expenditure on pension programs and other social security programs are very low in most Asian countries as compared with European and Latin American countries (Panel A, Table 35). In 1993, the percentage of public expenditure on social security programs was two percent in the Philippines and eight percent in Korea. In contrast, the percentages in other regions were far higher: 22 percent in the US, 31 percent in
Chile, and over 40 percent in most European countries. Likewise, the percentages of tax revenues contributed through participation in social security programs were generally lower in
Asia than elsewhere (Panel B, Table 35). In 1996, the figures were 1.4 percent in Thailand and
135 about 9 percent in Korea, as compared with 33 percent in the US and about 40 percent in Sweden and the Netherlands. Social security programs are becoming increasingly important in Asia, however. Social security’s proportion of tax revenue increased from 0.1 to 1.1 in Malaysia and from 0.7 to 8.8 in Korea between 1972 and 1996.
Table 35. Percentages of social security expenditure and tax revenue, selected years
The wide variation in the size of public pension programs is documented further by the percentages of old age benefit expenditure in GDP (Table 36). With the exception of Japan, most
Asian countries are spending less than two percent of GDP on pension benefits. In 1993, the percentages were only 0.22 percent in India, 0.14 percent in Korea, and zero percent in
Bangladesh, Thailand and Indonesia. Japan, Singapore and China were spending more than one percent of their GDP on old-age benefits. The percentages in North America and Europe were far higher: 6.3 percent in the US and 17.7 percent in Sweden, for example.
Table 36. Percentages of old-age benefit and other social security expenditure in GDP, circa
1993
The differences in spending on public pensions reflect coverage and average benefit levels, but also differences in the number of elderly and the maturity of pension programs.
Pension programs consume a larger share of public resources in the developed countries because a much larger proportion of their populations have reached retirement ages. Moreover, those programs were established many decades ago so that a high percentage of current retirees meet
136 eligibility requirements. As populations age and as programs mature in Asian countries, the share of public resources devoted to pensions will rise. Whether or not they will rise to the levels common in industrialized countries in the West will depend on the direction of public policy over the coming decades.
Reform of public support systems has received a great deal of attention in recent years especially in Latin America and the West. As the retired population has increased and public support systems have matured, the high costs of commitments made to the support of the elderly are becoming increasingly apparent. In some respects Asia is fortunate because aging is not as advanced as in the West, except in Japan, and because Asian public support programs are not as ambitious as those found in the West and many Latin American countries. Asia is at a critical juncture, however, because its populations are beginning to age and at a much more rapid pace than in the West. Further complicating matters is that some of the necessary pre-conditions for establishing sound public support systems, e.g., political stability and strong financial institutions, may not be met in some Asian countries. In the face of these circumstances, a strategy for establishing effective and sustainable public support system for the elderly is an urgent priority.
Should governments require universal participation in comprehensive programs?
Why should old-age security programs be mandatory? In general, redistributive goals cannot be achieved through voluntary programs. Hence, to the extent that old-age support programs are designed to reduce income inequality, mandatory participation is required.
Moreover, compulsory programs are adopted because individuals may be unable to adequately
137 provide for their retirement years, because of myopia, ignorance, or other reasons, when left to their own devices.
Although justified in some circumstances, mandatory programs also have their drawbacks. Mandatory programs may decrease welfare by requiring actions that individuals do not prefer. For example, compulsory saving programs may require individuals to accumulate more wealth, and consume less during their working years, than they would prefer. Mandatory programs may create inefficiency and slow economic growth by inducing behavioral changes.
Public support programs, for example, may reduce saving rates and lead to early retirement.
It is not always the case that transfer programs that target the elderly will reduce income inequality. The elderly, as a group, may have lower standards of living and higher rates of poverty than younger generations, but the extent to which this is true varies from country to country and over time. The economic status of the elderly depends on a variety of factors: (1) labor force and retirement policies and the labor force behavior of older workers; (2) the extent to which the elderly have accumulated wealth and, hence, the policies and systems that govern the financial sector; (3) the strength of the family support system; and, (4) unforeseen events that affect older individuals and the elderly as a group. For example, Asia’s recent financial crisis has almost certainly adversely affected the prospects for many elderly and older workers nearing retirement.
Even if the elderly are economically disadvantaged, old-age security programs may be a relatively inefficient method of redistributing income. Many programs provide benefits to the elderly irrespective of their economic status. Moreover, high-income individuals may receive greater lifetime benefits because of their higher life expectancy. Mandatory programs may not achieve their intended goals because of non-compliance, rent-seeking by strong political groups,
138 or behavioral changes such as reductions in the family support system (Valdés-Prieto). These problems are likely to be most severe when the program’s goals are not shared by a nation’s populace or when program participants are not confident in the ability of the government to achieve stated goals.
In light of these potential problems, governments should also explore alternative means of achieving the redistributive and life-cycle goals of old-age security programs. As discussed above, government policies or private employment practices may discriminate against older worker. Gender discrimination in various forms may adversely affect the economic circumstances of older women (who make up a disproportionately large share of the elderly).
Poorly developed financial institutions may interfere with the ability of workers to provide for their own old-age security. Directly addressing these problems can reduce the scale of mandatory old-age security programs and, consequently, their adverse effects.
Many Asian governments have, to this point, implemented public support systems on a more limited scale, in part, because of assessments about their capacity to efficiently operate such programs and, in part, because of assessments about the need for these programs. In many countries, a high proportion of the labor force is employed in the urban informal sector or the agricultural sector. There is a high incidence of casual workers and many workers who are underemployed. The lower paid self-employed, casual workers, domestic workers, agricultural workers and informal sector workers are often subject to low and irregular incomes and face uncertainty of about their day-to-day circumstances. Many of these workers may be unwilling to pay higher taxes in return for a promise of payments in the distant future. Moreover, there may be severe technical difficulties in collecting revenues in sectors where labor-turnover is high and
139 documentation is weak (Bailey, 1997a). As a consequence many pension systems throughout the region are restricted to public servants and those in the formal and larger private sectors.
Compliance problems have been serious in those cases where governments have attempted to expand old-age security programs too rapidly. Some developing countries have sought to extend coverage to some of the self-employed and even to employees such as domestic workers. The Philippines, for example, has accorded high priority to extending coverage to all private employees under age 60. Recent legislation requires that household help and self- employed workers be insured. In this instance, there is a probably a substantial gap between coverage under the law and coverage in practice (Bailey, 1997a). In practice, many schemes have experienced difficulty in ensuring compliance and registering employers as well as employees. Ambiguities in legislation have led to confusion about who should be covered.
Problems relating to the declaration of earnings have led to underpayment and late payments, and, in many instances, no payment at all.
The speed with which programs are extended may also be influenced by different assessments of the need for old-age support programs. In this regard, Asia faces conflicting prospects. Many Asian countries are experiencing population aging at a much faster rate than has occurred elsewhere. Consequently, countries like Indonesia, Thailand, Korea, and China will have to establish programs relatively quickly. However, even in the long-term government programs may prove to be a less important component in the old-age security of the elderly because of the persistence of the family support system and the high rates of saving found in many countries.
140 Financing: funded or pay-as-you-go?
Old-age security programs are funded in one of two ways. Under a “pay-as-you-go”
(PAYGO) system, current retirees are supported by transfers from current workers who will, in turn, be supported in old age by the next “generation” of workers. Current revenues cover current obligations, and there is no saving to pay future pensions. In a fully funded scheme, each generation finances its own retirement. Taxes paid by workers on their earning are accumulated, along with interest payments, in a retirement fund. The fund is then used to provide retirement benefits. Many programs are neither entirely funded nor entirely PAYGO. Some countries may accumulate trust funds that cover a portion of the future obligations. Historically, European countries and Latin American countries relied primarily on PAYGO systems, although recent reforms have promoted a shift to funded systems. Malaysia and Singapore have relied on funded systems, while South Korea, the Philippines and Egypt have programs that are partially funded and partially PAYGO (World Bank, 1994).
The distinction between PAYGO and funded financing applies equally well to private programs. Family-based transfer systems finance old age security by transferring income for family members who are of working age to family members who have retired. Thus, they are essentially PAYGO. When individuals accumulate wealth to provide for their own retirement, they are relying on a funded approach to financing old age security. It is not completely apparent which financing system is operating when multi-generation households are prevalent. Non- working elderly living in extended households may be receiving transfers from their children, or they may be receiving returns on financial wealth on investment they made in a household enterprise, or some combination of the two.
141 A key difference between PAYGO programs and funded programs is that funded programs do not transfer resources across generations. In contrast, the motivation for establishing an old-age security program is the view that elderly are in need of financial assistance.
Immediately raising the economic status of the elderly as a group can be accomplished only be transferring resources from those in the working ages. A PAYGO program may accomplish this, but a funded program will not. The impact of establishing a PAYGO system on the intergenerational distribution of income is uncertain because to some extent an increase in public transfers will reduce private transfers. The extent to which this occurs is not well-established
(World Bank, 1994).
In a mature old-age security program, a well-run funded system can support higher benefit levels than a PAYGO system because a funded system can achieve higher rates of return.
PAYGO financing can sustain an annual rate of return equal to the rate of economic growth (the rate of population growth plus the rate of productivity growth). The rate of return available from funded systems depends on the yield from the assets in which the retirement funds are invested.
On average, the sustainable returns are higher than those available from a PAYGO system depending on the risk associated with the investment strategy employed. A critical element of a successful funded system, however, is the management of the investment fund. Whether management is left in public or private hands, the potential for large-scale failures is a serious concern.
A potentially important advantage of funded systems is that they promote economic growth by raising rates of national saving. Pension funds, public and private, are an increasingly important source of capital. Countries with large mandatory funded systems, such as Singapore, have very high rates of public saving and national saving (Toh 1998, Baillu and Reisen 1997).
142 PAYGO systems may lead to lower rates of private saving by undermining the pension motive, although the empirical evidence on this issue is mixed (World Bank 1994).
Risk pooling and old-age support systems
Any support system for the elderly involves numerous risks. Workers may experience disability destroying their ability to provide for their old age security. Older adults are exposed to the risk of unusual longevity with the possibility that they will outlive the financial resources available for their support. Because of investment risk, some savers experience very high rates of return and others very low rates of return. Some savers may lose all of their savings. Workers who rely on public programs are subject to political and insolvency risk. The taxing ability of the government can decline or the political regime may change, with the new government repudiating the arrangements made by a previous government. Workers who depend on occupational retirement plans are subject to similar risks. In many countries, the pension plans of employers may not be funded and workers bear the risk of losing their pensions because of employer insolvency.
A variety of systems exist for pooling risks and reducing the exposure of retirees to the uncertainties. In the absence of other insurance arrangements, one of the chief disadvantages of family support systems is the limited ability of the family to pool risks. The family bears all of the costs when one of its members lives to an unusually old age, or when disability strikes, or when the family farm fails. One of the chief advantages of public systems is that some or all of these risks are pooled across all participants in the system. Most public defined benefit programs insure against longevity risk by providing a monthly benefit that lasts as long as the beneficiary survives. Of course, the lifetime benefit received by individuals who die at a young age are
143 correspondingly reduced. Some defined benefit and many defined contribution programs provide a lump-sum benefit at the time of retirement providing no insurance against longevity risk. In some countries, low cost annuities are available that allow participants in these programs to protect themselves against longevity risk. Most public, defined benefit programs also include a disability insurance component. Many defined contribution programs do not include a disability feature and, hence, alternative arrangements are necessary to insure against such risks.
The inherent difference between defined benefit and defined contribution programs is their treatment of investment risk. Defined contribution programs expose participants to investment risk and the benefits received by any cohort of retirees will be influenced, for example, by volatility in financial markets. Defined benefit programs spread the risks of volatile financial markets among all workers. In either approach, workers still bear the risks of uncertainties about their wage increases.
In many countries, the primary risk faced by a public programs is political and insolvency risk. PAYGO plans depend on the continued commitment of the state to honor obligations made by previous regimes. Given the political instability that exists in many countries in Asia and elsewhere in the developing world, these obligations may or may not be met. Funded systems are not immune from political risk. Pension funds can be diverted to other purposes by governments.
However, funded systems may provide more protection from political risk by clearly establishing the property rights of program participants. Some funded systems provide further protection by relying on multiple, private firms (Valdés-Prieto forthcoming, World Bank 1994).
144 Governance: centrally or privately managed?
In addition to questions of coverage and financing, governance problems are requiring reform in Asia. Issues range from institutional arrangement to management and operation. There are wide variety of institutional arrangements for the governance of social security in Asia, which lies in the spectrum between reliance on private insurance and direct administration by central government (Bailey, 1997b). In many Asian countries, the government is directly responsible for the administration of the social security scheme. The Social Security
Organization in Thailand, which is administered by divisions of the Ministry of Labor, is an example of such direct administration by central government. The Employees Provident Fund in
India is semi-autonomous with a Central Board of Trustees, but it is, in practice, subject to considerable control by the central government. In the Philippines, there are several autonomous social security institutions governed by their own supervisory board.
Should pension reserves be publicly or privately managed? Private management may be preferable because the profit motive and competition will lead to higher (risk-adjusted) returns and a broader range of choices. Public fund managers may be driven more by the interests of the political party in power than by the interest of beneficiaries. Publicly managed funds are usually invested in government securities or state enterprises often at below market interest rates (World
Bank, 1994). As governments obtain privileged access to large pension reserves, they may pursue large-scale public infrastructure projects in ways that are not subject to the scrutiny of open capital market. The danger is that choices are made without an explicit productivity comparison, without a market test based on interest rates and risk (World Bank, 1994).
Private administration may also bring important risks. Autonomous programs have been established to insulate them from political interference, but autonomy may lead to lower program
145 participation, especially low-income workers, if autonomous organizations are viewed as having less authority (Turner, 1997). Some countries with underdeveloped financial systems have been reluctant to allow private management of investment funds because of fears of fraud and incompetence. Effective regulation by the government may be difficult in many developing countries. The recent experience of Chilean reform suggests that reliance on private institutions does require a significant amount of administrative capabilities from the state, expressed in an ability to regulate financial intermediaries effectively (Valdés-Prieto, forthcoming).
Conclusion
Developing and implementing an effective strategy for dealing with old-age security issues is a difficult task. Countries must confront many complex technical issues. Decisions must be based on little more than educated guesses about very long-term changes in both demographic and economic characteristics. Capacities for implementing programs vary widely. Political considerations may in many instances establish additional road-blocks. Despite these difficulties, the rapid speed of aging in Asia requires a sound and comprehensive response.
Governments can take important, immediate steps that will benefit both broad development concerns and the elderly. Policies that increase labor market flexibility and remove barriers faced by older workers should be a priority. Developing a sound financial infrastructure will encourage higher rates of personal saving and allow for the emergence of the financial institutions needed for successful pension reform. The elimination of gender discrimination will allow women, the largest group of elderly, to better prepare for their later years in life.
Substituting saving for transfers is a central feature of many of the current reform efforts.
A system that emphasizes individual saving has many advantages. The sustainable rates of return
146 are higher than under PAYGO programs. Saving programs are less likely to encourage evasion or to encourage early retirement and they are less likely to be subverted by political pressure.
Moreover, saving is pro-growth. The World Bank summarizes the view as follows: “…many analysts would argue that saving would be growth-enhancing for most countries and for the world as a whole today. For these reasons, many capital-scarce countries are becoming increasingly interested in funded old age security arrangement” (World Bank, 1994).
A funded scheme has an important disadvantage: it is a weak tool for alleviating poverty among poor elderly. Thus, in developing an appropriate strategy, it is critical to assess the relative position of the old and other groups in the income distribution. Unless the elderly are disproportionately poor, a scheme that emphasizes redistribution may be avoidable. Moreover, to the extent that poverty among the elderly is a problem, an alternative is to address it through a comprehensive poverty alleviation program rather than an old-age security program.
A strategy many low-income countries have followed is to establish a PAYGO system.
The advantages of the approach are several. The immediate needs of the elderly are addressed.
Transfer programs are less complex to administer than funded programs. And given the very high support ratio typical of low-income countries, the fiscal burden is manageable. PAYGO programs, however, become increasingly unattractive, relative to funded programs, as countries develop better financial institutions and administrative capacity and their populations age.
Reform, however, is a difficult and costly process particularly if the existing PAYGO system provides substantial benefits and if population aging has advanced relatively far. In Latin
America several governments have reformed their programs and others have explored the possibility. In too many cases, however, the potential costs of the transition between the two
147 systems have not been carefully assessed, thereby risking failure in the reform process (Rofman, forthcoming).
Asia has an advantage over Latin America because large-scale PAYGO programs have been avoided. On the other hand, many Asian countries have barely begun to develop programs that will address the economic security of the elderly. Unfortunately, many Asian countries lack the legal structure, financial markets and administrative capacity, needed if comprehensive and effective programs are to be established. The family support system has served the needs of the elderly with considerable success. The traditional approach, however, will become an increasingly poor substitute for modern systems. Given the pace of aging in Asia, the region has a relatively short window of opportunity for implementing the programs and policies needed to protect the economic security of its older members.
148 Conclusions
Asia has entered a new demographic era. The number of children, those aged 0-14, reached a peak in the year 2000 after decades of rapid and sustained growth. As we move into the next century, the number of children may stabilize at the current level of around 300 – 350 million.
Or, if fertility rates drop to low levels, Asia could experience a period of sustained decline in the number of children.
During the next few decades, the adult population will continue to grow, older populations especially rapidly. Between 2000 and 2050, the 60-74 year old group is expected to increase at annual rate of 2.9% and the 75+ population at 3.4%. By 2050, 17% of Asia’s population will be 65 and older while the percentage under 15 will shrink to 19%. This is a remarkable change from the 1970s when the percentage under age 15 reached 40% of the total and the percentage 65 and older was only 4%.
Aging is a world-wide phenomenon, but the pace of aging is especially rapid in many
Asian countries which experienced rapid gains in life expectancy and precipitous declines in fertility. Japan, China, and several other East and Southeast Asian countries are expected to experience an especially rapid transition to an aged society. As a consequence, these countries will have less time to adjust to the social, economic, and political changes that accompany aging.
Some Asian countries are also experiencing population aging at much lower development levels than in the West. Indonesia’s old-age dependency ratio projected for 2050, for example, is a little higher than the ratio projected for the US in 2020. Even if Indonesia could achieve annual growth in per capita income of 4% per annum, a remote possibility, its per capita income in 2050 would reach only two-thirds of the current US level. Thus, Indonesia and other Asian countries may not be prepared for the challenges that population aging brings.
149 Forecasting demographic change for individual countries or for the region as a whole is, of course, a precarious exercise. Uncertainties abound. In some countries, even the current values of fertility, mortality, and population age-structure are a matter of dispute. Trends in fertility rates are difficult to predict. Will Pakistan experience rapid fertility decline? Will childbearing in Japan recovery from the low levels that persist there? Will fertility rise in China as it relaxes its one-child policy? How will fertility and other demographic variables be influenced by economic crisis in Southeast Asia?
Trends in mortality have been smoother and more predictable than trends in fertility.
After the end of World War II, the countries of Asia experienced rapid declines in infant and child mortality and steady increases in life expectancy. As mortality rates have reached low levels, the rate of increase has slowed but progress has continued to be steady. The recent mortality experience of countries in Africa and Eastern Europe cannot help but raise doubts about future mortality trends in Asia. HIV/AIDS presents the most immediate challenge in
Indochina and India. Thailand has made substantial progress in controlling the spread of the epidemic, but the same cannot be said for India, Burma, or Cambodia. The re-emergence of malaria and tuberculosis and the emergence of new infectious diseases may disrupt the steady progress that Asia has to this point enjoyed.
The implications of uncertainty for population age-structure in this study are explored using alternative fertility scenarios prepared by the United Nations Population Division and alternative mortality scenarios constructed by the authors. Taken together these scenarios reinforce general observations about population aging in Asia and the individual countries examined here. With some confidence we can anticipate rapid growth in elderly populations and higher old-age dependency ratios. Catastrophic events may lead to a different outcome, but the
150 more likely scenarios explored here all point to substantial population aging. The differences across scenarios are relatively modest in 2025, but much more substantial in 2050. Hence, long- range planning of the type needed in the design of old-age security programs, for example, must build in sufficient flexibility to accommodate uncertainty about future demographic trends.
Immigration patterns may also change in ways that influence population aging or mediate its impact. Historically, political crisis and war have precipitated mass migrations in Asia, and the future may hold more of the same. The region has not yet achieved a political stability that would rule out similar events. The possibilities are many: conflict between North and South
Korea, Taiwan and mainland China, or India and Pakistan; internal strife in the Philippines,
Indonesia, and Burma to name a few. The impact on population aging of large-scale refugee movements is uncertain, because the migration often involves entire communities rather than particular segments. This contrasts with economically motivated immigration typically dominated by young adults.
Most Asian countries have very restrictive immigration policies. Even in the face of severe labor shortages, Japan, for example, has admitted only a limited number of foreign workers on a temporary basis. Policies and attitudes towards immigration may change as populations age and labor force growth slows and, in some instances, turns negative. If so, then migration flows from younger (India, Pakistan, Bangladesh) to older (Japan, Korea, China) populations may attenuate the pace of aging in countries further along in their demographic transitions and speed the aging process in countries earlier in the transition.
As population aging proceeds in Asia, the region must deal with three broad sets of economic changes addressed in this report: changes in the size and basic character of the health
151 sector; the need to develop a much larger and comprehensive approach to old-age security; and, changes in macroeconomic conditions.
A thoroughly pessimistic view of the future of health care in Asia is that large cohorts of elderly will make ever-increasing demands on health systems, driving up costs, consuming larger shares of national income, burdening future generations of tax-payers, yielding little more than a few extra months or years of low-quality life. The available evidence does not support such a pessimistic view.
Health care expenditures as a percentage of national income are higher in OECD countries than in developing countries and they are rising over time. In part, the rise in health care expenditures is a reflection of the high priority attached to health by individuals and societies. More is being spent on health because it is so valued. The impact of aging on health care expenditure has not been firmly established. Longer-living populations are also healthier populations less in need of health care services. At the same time, longer-living populations are more heavily concentrated at older ages where health care needs are greater. In particular, the proportion of the population concentrated in the last year or two of life rises, contributing to higher health care expenditures. In some countries, the system of health care finance is a major contributor to rising health care expenditure. Direct and indirect subsidization of health care is leading to inefficiency in the production of health care services, over-use of services by consumers, and behavior by individuals that is subject to higher health risks, e.g., smoking.
Population aging is leading to a shift in the ways in which health care resources are being used. Expenditures on reproductive health, public health measures, and the treatment of infectious diseases are becoming relatively less important. Expenditures on the treatment of acute and chronic conditions, such as, cardiovascular diseases and cancer are becoming more
152 important. These shifts clearly influence the kinds of services the health care industry provides, but the changes also have important implications for health care policy. When the key health problems are the prevalence infectious diseases or an unsafe water supply, the existence of externalities insures that too few resources will be devoted to health care in the absence of active government involvement. But when acute and chronic conditions come to dominate the health care agenda, government involvement is more likely to result in excessive resources devoted to health care, in general, or to particular forms of health care. The key challenge for Asian countries over the coming decades will be to continually reform their health care systems in congruence with the changes in health care needs.
In Asia as in most societies, supporting children is primarily a family responsibility.
Governments may provide or subsidize education and health care, but the principle costs of rearing children are borne by parents and other family members. Supporting the elderly also has been primarily a family responsibility, but the family support system is evolving and, in some respects, is in decline. In some countries, elderly are already relying less on their children for their material needs and more on a combination of wealth accumulated during their working years and public transfer programs financed by taxing workers. These changes are well underway in some Asian countries and on the horizon in others.
The family has been an effective provider of old-age support, and will continue to serve an important role in the future. Modern institutions are a poor substitute for the personal care and attention that families provide. But the family also has its limitations. The family is poorly equipped to deal with many of the risks associated with aging. Family “contracts” are largely unenforceable exposing the elderly to the danger that their children will default on their commitments. Family support systems may provide a low rate of return as compared with
153 alternatives. These limitations are becoming increasingly apparent as social and economic development undermines traditional values and as elderly populations grow relative to the population expected to provide support.
There are two broad challenges in the old-age support area. One is to provide an enabling environment that will allow older adults to maintain as much economic independence as is possible. Three areas appear to be particularly important. First, governments can ban employment policies that encourage early retirement and reduce job opportunities and job mobility for older workers. Second, governments can establish a regulatory environment and, in other ways, encourage the development of the financial sector that will allow workers to save for their retirement. Third, governments can eliminate discriminatory practices, including those governing education, employment, property ownership, and inheritance, which undermine the economic independence of older women.
The second challenge is to design a public system of old-age support that promotes economic self-sufficiency, satisfies distributional objectives, and insures the elderly against risks, such as, longevity risk. Governments must be certain that they have the ability to operate programs they choose to implement and, where appropriate, should rely on public-private partnership insuring an efficient and equitable support system. Many Asian governments have an advantage. They have not adopted large-scale programs that are costly and difficult to reform. However, the development of comprehensive and sound programs has become an urgent matter.
The success with which countries meet challenges in the health and old-age security area will depend, in large part, on the economic performance of their economies. Pessimistic views about the impact of aging on economic growth abound and with some reason. In the coming
154 decades, labor force growth will slow and, in some countries, turn negative. The workforce will become older and more experienced, but possibly less mobile and innovative. Rates of saving and investment may decline significantly from the high levels that currently fuel much of the region’s economic growth.
Slower economic growth may well be inevitable, but little cause for concern if countries achieve standards of living comparable to the levels now found in Japan, Singapore, and Hong
Kong. A growing body of evidence suggests that increases in life expectancy have made an important contribution to economic growth by encouraging investment in human capital and higher rates of saving. Thus, aging societies may be poorly endowed in labor quantity, but well- endowed in physical and human capital. This optimistic outcome will only be realized, however, if countries adopt policies that encourage saving, investment in education, the development of well-functioning financial institutions, a favorable investment environment, and integration into the global economy.
155
Figure 1. Asia's age transition
100
Elderly (65+) 90
80 Children (0-14)
70
60
50 Working age (15-64)
40
Percentage of total population 30
20
10
0 1950 1960 1970 1980 1990 2000 2010 2020 2030 2040 2050 Figure 2. Age pyramid for Asia 80+
75-79
70-74
65-69
60-64
55-59
50-54
45-49
40-44
35-39 Age group 30-34
25-29
20-24
15-19
10-14
5-9
0-4
200,000 150,000 100,000 50,000 0 50,000 100,000 150,000 200,000
80+ 2025 75-79 70-74 65-69 60-64 55-59 50-54 45-49 40-44
Age group 35-39 30-34 25-29 20-24 15-19 10-14 5-9 0-4
200,000 150,000 100,000 50,000 0 50,000 100,000 150,000 200,000
80+
75-79
70-74
65-69
60-64
55-59
50-54
45-49
40-44
35-39
30-34
25-29
20-24
15-19
10-14
5-9
0-4
200,000 150,000 100,000 50,000 0 50,000 100,000 150,000 200,000 Figure 3. Population of Asia fifteen year age groups, 1950-2050
1200000
0-14 1000000 15-29 30-44 45-59 800000
60-74
600000 Population (1000s)
400000 75+
200000
0 1950 1965 1980 1995 2010 2025 2040 Figure 4: Male labor force of Asia ten year age group 1950-2050 400000
350000
300000 10-19 20-29 250000 30-39 40-49 200000 50-59 60+ 150000
Labor Force (1000s) 100000
50000
0 1950 1960 1970 1980 1990 2000 2010 2020 2030 2040 2050
Source: ILO, 1996. UN, 1999. Calculations by authors. Figure 5. Female labor force of Asia Ten year age group 1950-2050 300000
250000
200000 10-19 20-29 30-39 40-49 150000 50-59 60+ 100000 Labor Force (1000s) 50000
0 1950 1960 1970 1980 1990 2000 2010 2020 2030 2040 2050
Source: ILO, 1996. UN, 1999. Calculations by authors. Figure 6. Male labor force participation rates of Asia by age group
1.2
1
0.8
0.6 1950 1970 1990 0.4 2010
0.2
0 10-1415-1920-2425-2930-3435-3940-4445-4950-5455-5960-6465 + Source: ILO, 1996. Figure 7. Female labor force participation rates of Asia by age group
0.8
0.7
0.6
0.5
0.4 1950 1970 0.3 1990 2010 0.2
0.1
0 10-1415-1920-2425-2930-3435-3940-4445-4950-5455-5960-6465 + Source: ILO, 1996. Figure 8. Male labor force transition in Asia
100% 60+ 90%
80% 50-59
70% 40-49 60%
50% 30-39
40%
30% 20-29
20%
10% 10-19
0% 1950 1960 1970 1980 1990 2000 2010 2020 2030 2040 2050
Source: ILO, 1996. UN, 1999. Calculations by authors. Figure 9. Female labor force transition of Asia
100% 60+ 90% 50-59 80%
70% 40-49
60% 30-39 50%
40% 20-29 30%
20%
10% 10-19
0% 1950 1960 1970 1980 1990 2000 2010 2020 2030 2040 2050
Source: ILO, 1996. UN, 1999. Calculations by authors. Figure 10A. Projections of proportion widowed, South Korea 1
Females 2000 0.8 2025 0.6 2050 0.4
0.2
0 1 35-39 40-44 45-49 50-54 55-59 60-64 65-69 70-74 75-79 80-84 85+
Males 0.8
0.6 2000 0.4 2025
0.2 2050 0 35-3940-4445-49 50-5455-5960-6465-6970-74 75-7980-84 85+ Figure 10B. Projections of proportion widowed, The Philippines 1
Females 0.8
0.6 2000
2025 0.4
0.2 2050 0 1 35-39 40-44 45-49 50-54 55-59 60-64 65-69 70-74 75+
Males 0.8
0.6
0.4 2000
0.2 2025 0 2050 35-39 40-44 45-49 50-54 55-59 60-64 65-69 70-74 75+ Figure 10C. Projections of proportion widowed, Indonesia 1 Females 0.8 2000
0.6 2025
2050 0.4
0.2
0 1 35-39 40-44 45-49 50-54 55-59 60-64 65-69 70-74 75+ Males 0.8
0.6
0.4
2000 0.2 2025
2050 0 35-39 40-44 45-49 50-54 55-59 60-64 65-69 70-74 75+ Figure 10C. Projections of proportion widowed, Indonesia 1
Females 0.8
0.6 2000
0.4 2020
0.2 2050 0 1 35-39 40-44 45-49 50-54 55-59 60-64 65-69 70+ Males 0.8
0.6
0.4
0.2 2000
2020 0 2050 35-39 40-44 45-49 50-54 55-59 60-64 65-69 70+ Figure 11. Global fertility transition UN 1998 revision, medium variant
0.7
0.6
0.5 median
0.4 90th percentile 10th percentile 0.3
0.2
0.1
TFR decline (births per five year period) 0
-0.1
-0.2 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 Total fertility rate Figure 12. Analysis of TFR projection comparison of medium variant to global historical pattern
1.6
1.4
1.2
1
0.8 90th percentile
0.6 UN medium variant
0.4 median
TFR decline (births per five year period) 0.2
10th percentile 0
-0.2 1 2 3 4 5 6 7 Total fertility rate Figure 13. Analysis of TFR projection comparison of high and low variants to global historical pattern
1.6
1.4
1.2
1 90th percentile
0.8
UN low variant 0.6
0.4
UN high variant
TFR decline (births per five year period) 0.2 median 10th percentile 0
-0.2 1 2 3 4 5 6 7 Total fertility rate Figure 14. Global mortality transition UN population projections
5
4 90th percentile
3 Median
2
1
Change in e0 per five year period 10th percentile
0
-1 40 45 50 55 60 65 70 75 80 85 Life expectancy at birth Figure 15. Life expectancy transition 1950-1995 5
4
90th percentile 3
2 Median
1 10th percentile Increase per five-year period 0 30 35 40 45 50 55 60 65 70 75 80
-1
-2 Life expectancy at birth Note: Countries with 1990 population exceeding 2 million. Figure 16. Comparison of UN mortality projections to historical data
3
1950-95 Historical 2.5
2
1985-95 Historical
1.5
1 Change in e0 per five-year period
UN Projections 0.5
0 50 55 60 65 70 75 80 85 Life expectancy at birth Figure 17. Economic support ratio, Southeast Asia
0.72
0.7
0.68
0.66
0.64
Workers per equivalent consumer 1950 1960 1970 1980 1990 2000 2010 2020 2030 2040 2050
0.3 0.2 0.1 0 -0.1 -0.2 -0.3
Annual rate of growth -0.4 -0.5 1950s 1960s 1970s 1980s 1990s 2000s 2010s 2020s 2030s 2040s -0.40 0.18 -0.20 Figure 18. Growth Rates (%) of Economic Support Ratios by Countries
Japan 0.60 -0.26
South Korea -0.85 0.42 -0.41
The Philippines 0.14 -0.14
Thailand -0.26 0.25 -0.37
Indonesia -0.47 0.36 -0.20
Bangladesh -0.23 0.26 -0.17
India -0.20 0.15 -0.18
Egypt -0.10 0.35 -0.28
1950s 1960s 1970s 1980s 1990s 2000s 2010s 2020s 2030s 2040s
Growth Rates (%) of Economic Support Ratios by Regions
Asia -0.17 0.10 -0.23
East Asia -0.17 0.22 -0.56
South East Asia -0.40 0.18 -0.20
South Asia -0.19 0.30 -0.17
1950s 1960s 1970s 1980s 1990s 2000s 2010s 2020s 2030s 2040s Figure 19. Wealth/Output : Taiwan, US, & France, 1800-2100
Taiwan US France
67 interest rate= 3.0%, prod. rate=1.5%
France US Taiwan ratio 345 12
1800 1850 1900 1950 2000 2050 2100
Date Figure 20. Savings Rate: Taiwan, US, & France, 1800-2100 25
interest rate= 3.0%, prod. rate=1.5%
Taiwan
20 Taiwan US France
15 United States Percentage 10 5
France 0
1800 1850 1900 1950 2000 2050 2100
Date Figure 21. Investment, 1960 vs. 1990 Slow and fast transition populations
50
45
40 Japan S Korea 35 Singapore
Indonesia 30 Thailand
25 Taiwan 20 Philippines India 15
10 Pakistan Investment as a percentage of GDP 1990
5 Egypt Bangladesh 0 0 5 10 15 20 25 30 35 40 45 50 Investment as a percentage of GDP 1960 Source: Penn World Tables Figure 22. Gross National Saving Japan, 1885-1997, 2025 forecasts
45
40
35
30
25
20
15 Saving/Income (percentage)
10
5
0 1860 1880 1900 1920 1940 1960 1980 2000 2020 2040
Source: Mason and Ogawa, forthcoming. Figure 23. Rates of total expenditure on health by GDP
9
Japan
6
% of GDP Korea
3
0 1960 1965 1970 1975 1980 1985 1990 1995 Figure 24. Burden of death
0.300
0.250 Japan South Korea India 0.200
0.150
0.100 Non-infant deaths/population 15-64
0.050
0.000 19501955196019651970197519801985199019952000200520102015202020252030203520402045 Figure 25. Retirement hazard rates, Asian men 1950-2010
0.5
0.45
0.4
0.35
60-64 to 65+ 0.3 55-59 to 60-64 50-54 to 55-59 0.25 45-49 to 50-54 40-44 to 45-49 Hazard rate 0.2
0.15
0.1
0.05
0 1950 1960 1970 1980 1990 2000 2010 Source: ILO (1996), UN (1998). Figure 26: Relationship between activity rates of elderly men and the percentage of employment in agriculture sector,1990
90
80
70
60
50
40
30
20 Activity rates of men aged 65 + 10
0 0 10 20 30 40 50 60 70 80 90 100 Percentage of employemnt in agriculture sector Source: ILO (1996). References
Afzal, Mohammad. 2000. “Family Structure and the Elderly in Pakistan.” The Family and Older Persons in Bangladesh, Pakistan and Sri Lanka, Economic and Social Commission for Asia and the Pacific (ESCAP) Asian Population Studies Series No. 151. New York: United Nations.
Anderson, Gerard F., and Peter Sotir Hussey. 2000. “Population Aging: A Comparison Among Industrialized Countries.” Health Affairs 19(3) (May/June): 191-203.
Andreano, Ralph, and Thomas Helminiak. 1987. “The Role of the Private Sector in Health Care in the Developing Countries.” In Health Care Financing. Manila: Asian Development Bank, Economic Development Institute of the World Bank, and the East-West Center: 237-99.
Andrews, Gary R., and Monique M. Hennink. 1992. “The Circumstances and Contributions of Older Persons in Three Asian Countries: Preliminary Results of a Cross-national Study.” Asia- Pacific Population Journal 7(3): 127-46.
Arai, Yumiko, and Naoki Ikegami. 1998. “Health Care Systems in Transition II. Japan, Part I. An Overview of the Japanese Health Care System.” Journal of Public Health Medicine 20(1): 29-33.
Arnott, Richard J., and Joseph E. Stiglitz. 1988. “The Basic Analytics of Moral Hazard.” Scandinavian Journal of Economics 90(3): 383-413.
Arnott, Richard J., and Joseph E. Stiglitz. 1990. “The Welfare Economics of Moral Hazard.” In Henri Loubergé (ed.), Risk, Information and Insurance, pp. 91-121. Boston: Kluwer.
Atinc, Tamar Manuelyan, and Michael Walton. 1998. Social Consequences of the East Asian Financial Crisis. Washington D.C.: World Bank.
Auer, Peter, and Mariàngels Fortuny. 2000. Ageing of the Labour Force in OECD Countries: Economic and Social Consequences, Working paper 2000-2. Geneva: ILO.
Averill, Richard F. 1991. “Development.” In Robert B. Fetter (ed.), DRGs: Their Design and Development, Chapter 2, pp. 28-56. Ann Arbor, Michigan: Health Administration Press.
Bailey, Clive. 1997a. “Coverage under Social Security Pension Schemes.” Tripartite regional consultation with Asian countries, Bangkok, October 1997, ILO Social Security Department.
Bailey, Clive. 1997b. “Governance Issues.” Tripartite regional consultation with Asian countries, Bangkok, October 1997, ILO Social Security Department.
Baillu, Jeanine, and Helmut Reisen. 1997. “Maintaining Prosperity in an Ageing Society: The OECD Study on the Policy Implications of Ageing.” Ageing Working Papers AWP 5.1. Paris: OECD.
156 Behrman, Jere. 1997. “The Effect of Structural Adjustment on Food Policy and Nutrition.” In Georges Tapinos et al. (eds.), Demographic Responses to Economic Adjustment in Latin America. Oxford: Claredon Press.
Berkovec, James, and Steven Stern. 1991. “Job Exit Behavior of Older Men.” Econometrica 59(1): 189-210.
Berman, Peter A. 1997. “National Accounts in Developing Countries: Appropriate Methods and Health Economics 6(1): 11-30.
Berman, Peter A. 1998. “Rethinking Health Care Systems: Private Health Care Provision in World Development 26(8): 1463-79.
Blomqvist, Å. G., and R. A. L. Carter. 1997. “Is Health Care Really a Luxury?” Journal of Health Economics 16: 207-29
Blöndal S., and S. Scarpetta. 1998. “Retire Early, Stay at Work?” OECD Observer, no. 212.
Bloom, David E., and Jeffrey G. Williamson. 1998. “Demographic Transitions and Economic World Bank Economic Review 12(3): 419–56.
Bos, Edward, Vivian Hon, Akiko Maeda, Gnanaraj Chellaraj, and Alexander Preker. 1999. Health, Nutrition, and Population Indicators: A Statistical Handbook. Washington, D.C.: The World Bank.
Boserup, Ester. 1965. The Conditions of Agricultural Growth. Chicago: Aldine.
Boserup, Ester. 1981. Population and Technological Change: A Study of Long-term Trends. Chicago: University of Chicago Press.
Bossert, Thomas, Joel Beauvais, and Diana Bowser. 2000. Decentralization of Health Systems: Preliminary Review of Four Country Case Studies. Major Applied Research 6, Technical Report No. 1. Bethesda, MD: Partnerships for Health Reform Project, Abt Associates Inc.
Bovenberg, Lans, and Anja van der Linden. 1997. “Pension Policies and the Aging Society.” The OECD Observer, no. 205, pp. 10-14.
Bravo, Jorge. 1997. “Demographic Consequences of Structural Adjustment in Chile.” In Georges Tapinos et al. (eds.), Demographic Responses to Economic Adjustment in Latin America. Oxford: Claredon Press.
Busse, Reinhard, and Friedrich Wilhelm Schwartz. 1997. “The Philippines’ National Health Insurance Act: A German Perspective.” International Journal of Health Planning and Management 12(2): 131-48.
157 Campbell, John Creighton, and Naoki Ikegami. 2000. “Long-Term Care Insurance Comes to Health Affairs 19(3) (May/June): 26-39.
Casterline, J. B., L. Williams, A. Hermalin, M. C. Chang, N. Chayovan, P. Cheung, L. Domingo, J. Knodel, and M. Ofstedal. 1991. “Differences in the Living Arrangements of the Elderly in Four Asian Countries: The Interplay of Constraints and Preferences.” Comparative Study of the Elderly in Asia, Research Reports No. 91-10. Ann Arbor: Population Studies Center, University of Michigan.
Chan, A. 1997. “An Overview of the Living Arrangements and Social Support Exchanges of Asia-Pacific Population Journal 12(4): 35-50.
Cutler, David M., James Poterba, Louise M. Sheiner, and Lawrence H. Summers, 1990. “An Aging Society: Opportunity or Challenge?” Brookings Papers on Economic Activity 1, pp. 1-71.
Deardorff, Alan V. 1987. “Trade and Capital Mobility in a World of Diverging Populations.” In D. Gale Johnson and Ronald D. Lee (eds.), Population Growth and Economic Development: Issues and Evidence, pp. 561–88. Madison, WI: University of Wisconsin Press.
Deaton, Angus S., and Christina H. Paxson. 2000. “Growth, Demographic Structure, and National Saving in Taiwan.” Population and Development Review, forthcoming.
Department of Planning, Ministry of Health and Data for Decision Planning Project, Harvard School of Public Health. 1998. Health Care Utilization and Expenditures in the Arab Republic of Egypt 1994-95. Boston, MA: Harvard School of Public Health.
Domingo, Lita J., and John B. Casterline. 1992. “Living Arrangements of the Filipino Elderly.” Asia-Pacific Population Journal 7(3): 63-88.
Feeney, Griffith, and Andrew Mason. 1998. “Population in East Asia.” East-West Center Working Papers, Population Series, No. 88-2, Honolulu, December.
Feldstein, Martin, and Bernard Friedman. 1977. “Tax Subsidies, the Rational Demand for Insurance and the Health Care Crisis.” Journal of Public Economics 7(2): 155-78.
Fetter, Robert B. 1991. “Background.” In Robert B. Fetter (ed.), DRGs: Their Design and Development, Chapter 1, pp. 3-27. Ann Arbor, Michigan: Health Administration Press.
Filmer, Deon, Jeffrey Hammer, and Lant Pritchett. 1997. Health Policy in Poor Countries: Weak Links in the Chain. Policy Research Working Paper No. 1874. Washington, D.C.: The World Bank.
Fogel, Robert W. 1999. “Catching Up with the Economy.” American Economic Review 89(1) (Mar.): 1-21.
158 Gerdtham, Ulf-G., Jes Søgaard, Fedrik Andersson, and Bengt Jönsson. 1992. “An Econometric Analysis of Health Expenditure: A Cross-section Study of OECD Countries.” Journal of Health Economics 11(1) (May): 63-84.
Getzen, Thomas E. 1992. “Population Aging and the Growth of Health Expenditures.” Journal of Gerontology: Social Science 47(3): 98-104.
Getzen, Thomas E. 2000. “Health Care Is an Individual Necessity and a National Luxury: Applying Multilevel Decision Models to the Analysis of Health Care Expenditures.” Journal of Health Economics 19(2) (Mar.): 259-70.
Gillion, Colin. 1997. “Overview.” Tripartite regional consultation with Asian countries, Bangkok, October 1997, ILO Social Security Department.
Greenspan, Nancy T., and Ronald J. Vogel. 1980. “Taxation and its Effect upon Public and Private Health Insurance and Medical Demand.” Health Care Financing Review 1(4): 39-45.
Gustman, Alan L., and Thomas Steinmeier. 1994. “Employer Provided Health Insurance and Retirement Behavior.” Industrial and Labor Relations Review 48(1): 124-40.
Gustman, Alan. L., Olivia S. Mitchell, Thomas L. Steinmeier, and Andrew Samwick. 1997. “Pension and Social Security Wealth in the Health and Retirement Survey.” NBER Working Paper Series, no. 5912.
Hadley, Jack. 1991. “Theoretical and Empirical Foundations of the Resource-based Relative Frech III (ed.), Regulating Doctors' Fees: Competition, Benefits, and Controls under Medicare, pp. 97-125. Washington, D.C.: American Enterprise Institute.
Hayami, Yujiro, and Vernon W. Ruttan. 1987. “Population Growth and Agricultural In D. Gale Johnson and Ronald D. Lee (eds.). Population Growth and Economic Development: Issues and Evidence, pp. 57-104. Madison, WI: University of Wisconsin Press.
HCFA. 1998. Health Care Financing Review: Medicare and Medicaid Statistical Supplement, 1998. Baltimore, MD: U.S. Department of Health and Human Services, Health Care Financing Administration, Office of Strategic Planning.
Hermalin, Albert I. 1997. “Drawing Policy Lessons for Asia from Research on Ageing.” Asia- Pacific Population Journal 12(4): 89-102.
Hitiris, Theo, and John Posnett. 1992. “The Determinants and Effects of Expenditures in Developed Countries.” Journal of Health Economics 11(2): 173-81.
Hsiao, William C. et al. 1988a. “Resource-based Relative Values: An Overview.” Journal of the American Medical Association 260(16): 2347-53.
159 Hsiao, William C. et al. 1988b. “Results Potential Effects, and Implementation Issues of the Resource-based Relative Value Scale.” Journal of the American Medical Association 260(16): 2429-38.
Hsiao, William C., and Daniel L. Dunn. 1991. “The Resource-based Relative Value Scale for Pricing Physicians’ Services.” In H. E. Frech III (ed.), Regulating Doctors' Fees: Competition, Benefits, and Controls under Medicare, pp. 221-36. Washington, D.C.: American Enterprise Institute.
Hurd, Michael D. 1990. “Research on the Elderly: Economic Status, Retirement, and Journal of Economic Literature 118: 565-637.
Hurd, Michael D. 1996. “The Effect of Labor Market Rigidities on the Labor Force Behavior of Older Workers.” In David A. Wise (ed.), Advances in the Economics of Aging, pp. 11-60, NBER Project Report. Chicago: The University of Chicago Press.
Ikegami, Naoki. 1991. “Japanese Health Care: Low Cost through Regulated Fees.” Health Affairs 10(3) (Fall): 87-109.
Ikegami, Naoki. 1997. “Public Long-term Care Insurance in Japan.” JAMA 278 (16) (Oct. 22- 29): 1310-14.
Ikegami, Naoki. 1998. “Growing Old in Japan.” Age and Ageing 27: 277-78.
Ikegami, Naoki. 1999. “The Rationales for Establishing Independent Long-term Care.” In Aging and Health: A Global Challenge for the Twenty-First Century —Proceedings of WHO Symposium Kobe, 10-13 November 1998, pp. 373-76. Kobe, Japan: World Health Organization, Center for Health Development.
Ikegami, Naoki, and John Creighton Campbell. 1999. “Health Care Reform in Japan: The Health Affairs 18(3) (May/June): 56-75.
ILO, various years. Social Security Department Database.
ILO, various years. World Employment Report. Geneva: ILO.
ILO. 1996. Economically Active Population 1950-2010. Geneva: ILO.
Jacobzone, S., E. Cambois, E. Chaplain, and J. M. Robine. 1998. The Health of Older Persons in OECD Countries: Is It Improving Fast Enough to Compensate for Population Ageing? Labour Market and Social Policy Occasional Papers No. 37. Paris: OECD.
Jitapunkul, Sutthichai, Mongkol Na Songkhla, Napaporn Chayovan, Aroon Chirawatkul, Chanpen Choprapawon, Yongyout Kachondham, and Supakorn Buasai. 1999. “A National Survey of Health-service Use in Thai Elders.” Age and Ageing 28 (Jan.): 67-71.
160 Johnson, D. Gale, and Ronald D. Lee (eds.). 1987. Population Growth and Economic Development: Issues and Evidence. Madison, WI: University of Wisconsin Press.
Jowett, Mathew. 1999. “Bucking the Trend? Health Care Expenditures in Low-income Countries International Journal of Health Planning and Management 14(4): 269-85.
Keilman, Nico. 1999. “How Accurate Are the United Nations World Population Projections?” In Wolfgang Lutz, James W. Vaupel, and Dennis A. Ahlburg (eds.), Frontiers of Population Forecasting, a supplement to Population and Development Review (24): 15-41.
Kelley, Allen C. 1988. “Economic Consequences of Population Change in the Third World.” Journal of Economic Literature 26(4): 1685–1728.
Kelley, Allen C., and Robert M. Schmidt. 1996. “Saving, Dependency and Development.” Journal of Population Economics 9: 365-86.
Kim, Cheong-Seok. 1997. “Who Should Take Responsibility for Supporting the Elderly: Views Research Reports No. 97-398. Ann Arbor: Population Studies Center, University of Michigan.
Knodel, John, and Napaporn Chayovan. 1997. “Family Support and Living Arrangements of Asia-Pacific Population Journal 12(4): 51-68.
Knodel, John, and Nibhon Debavalya. 1997. “Living Arrangements and Support among the Elderly on South-East Asia: An Introduction.” Asia-Pacific Population Journal 12(4): 5-16.
Kochar, Anjini. 1999. “Evaluating Familial Support for the Elderly: The Intrahousehold Allocation of Medical Expenditures in Rural Pakistan.” Economic Development and Cultural Change 47 (Apr.): 621-56.
Koinuma, N. 1999. “Long-term Nursing Care Insurance in Preparation for the Super-aged Aging and Health: A Global Challenge for the Twenty-First Century —Proceedings of WHO Symposium Kobe, 10-13 November 1998, pp. 253-57. Kobe, Japan: World Health Organization, Center for Health Development.
Lam, David. 1997. “Demographic Variables and Income Inequality.” In M. R. Rosenzweig and O. Stark (eds.), Handbook of Population and Family Economics, pp. 1015–59. Amsterdam: Elsevier Science B.V.
Lazear, Edward P. 1979. “Why is There Mandatory Retirement?” Journal of Political Economy 87: 1261-64.
Lazear, Edward P. 1985. “Incentive Effects of Pensions.” In David A. Wise (ed.), Pension, Labor, and Individual Choice, pp. 253-82, NBER Project Report. Chicago: The University of Chicago Press.
161 Lee, Hung-Tak. 1998. Family Welfare and Reproductive Health: The Korean Experience, KIHASA/UNFPA Seminar on Population and Development Policies in Low Fertility Countries. Seoul: KIHASA.
Lee, Ronald D. 1993. “Modeling and Forecasting the Time Series of US Fertility: Age Patterns, International Journal of Forecasting 9: 187-202.
Lee, Ronald D. 1999. “Probabilistic Approaches to Population Forecasting.” In Wolfgang Lutz, James W. Vaupel, and Dennis A. Ahlburg (eds.), Frontiers of Population Forecasting, a supplement to Population and Development Review (24): 156-90.
Lee, Ronald, Andrew Mason, and Tim Miller. 2000. “From Transfers to Individual Responsibility: Implications for Savings and Capital Accumulation in Taiwan and the United States,” a paper presented to the Population Association of America Annual Meetings, San Diego, CA. March.
Lee, Ronald, Andrew Mason, and Timothy Miller, forthcoming a. “Saving, Wealth, and the Demographic Transition in East Asia.” East-West Center Working Papers: Population Series, 88- 7, Honolulu, HI.
Lee, Ronald, Andrew Mason, and Timothy Miller, forthcoming b. “Saving, Wealth, and Population.” Paper presented at the Symposium on Population Change and Economic Development, Bellagio Center, Lake Como, Italy, November 1998.
Lee, Ronald, Andrew Mason, and Timothy Miller, forthcoming c. “Life Cycle Saving and the Demographic Transition in East Asia. In Cyrus Chu and Ronald Lee, eds., Population and Economic Change in East Asia, a special issue of Population and Development Review.
Luft, Harold S. 1981. Health Maintenance Organizations: Dimensions of Performance. New York: Wiley.
Lumsdaine, Robin L., and Olivia S. Mitchell. 1999. “New Developments in the Economic In Orley C. Ashenfelter and David Card (eds.), Handbook of Labor Economics, Vol. 3C, Elsevier, pp. 3261-3307.
Lumsdaine, Robin R., James H. Stock, and David A. Wise. 1994. “Pension Plan Provisions and Retirement: Men and Women, Medicare, and Models.” In David A. Wise (ed.), Studies in the Economics of Aging, pp. 183-222, NBER Project Report. Chicago: The University of Chicago Press.
Lumsdaine, Robin R., James H. Stock, and David A. Wise. 1996. “Why are Retirement Rates So High at Age 65?” In David A. Wise (ed.), Advances in the Economics of Aging, pp. 61-82, NBER Project Report. Chicago: The University of Chicago Press.
Lumsdaine, Robin R., James H. Stock, and David A. Wise. 1997. “Retirement Incentives: The Interaction between Employer-Provided Pensions, Social Security and Retiree Health Benefits.”
162 In Michael D. Hurd and Naohiro Yashiro (eds.), The Economic Effects of Aging in the United States and Japan, pp. 261-93, NBER Project Report. Chicago: The University of Chicago Press.
Lutz, Wolfgang, Warren C. Sanderson, and Sergei Scherbox. 1999. “Expert-Based Probabilistic Population Projections.” In Wolfgang Lutz, James W. Vaupel, and Dennis A. Ahlburg (eds.), Frontiers of Population Forecasting, a supplement to Population and Development Review (24): 139-55.
Manning, Willard G., Joseph P. Newhouse, Naihua Duan, Emmett B. Keeler, Arleen Leibowitz, and M. Susan Marquis. 1987. “Health Insurance and the Demand for Medical Care: Evidence American Economic Review 77(3): 251-77.
Marshall, John M. 1976. “Moral Hazard.” American Economic Review 66(5): 880-90.
Martin, Linda G., and Noriko O. Tsuya. 1994. “Middle-Aged Japanese and Their Parents: Coresidence and Contact.” In Lee-Jay Cho and Moto Yada (eds.), Tradition and Change in the Asian Family, pp. 153-78. Honolulu: East-West Center.
Mason, Andrew. 1987. “National Saving Rates and Population Growth: A New Model and New Evidence.” In D. Gale Johnson and Ronald D. Lee (eds.), Population Growth and Economic Development: Issues and Evidence, pp. 523-60. Madison, WI: University of Wisconsin Press.
Mason, Andrew. 1988. “Saving, Economic Growth and Demographic Change.” Population and Development Review 14 (Mar.): 113-44.
Mason, Andrew. 1997. “The Response of Fertility and Mortality to Economic Crisis and Structural Adjustment Policy During the 1980s: A Review.” In Georges Tapinos et al. (eds.), Demographic Responses to Economic Adjustment in Latin America. Oxford: Clarendon Press.
Mason, Andrew. 1999a. “Population Change in East Asia.” In Andrew Mason, Thomas Merrick, and R. Paul Shaw (eds.), Population Economics, Demographic Transition, and Development: Research and Policy Implications, World Bank Institute Working Papers. Washington, D.C.: The World Bank.
Mason, Andrew. 1999b. “Population and Economic Growth in East Asia” East-West Center Working Papers, Population Series, No. 99-25, Honolulu (November).
Mason, Andrew, and Tim Miller. 2000. “Dynasties, Intergenerational Transfers, and Lifecycle Tapinos (eds.), Sharing the Wealth: Demographic Change and Economic Transfers between Generations.
Mason, Andrew, and Naohiro Ogawa, forthcoming. “Population, Saving, Labor Force, and Japan’s Future.” In Magnus Blomström, Byron Gangnes, and Sumner La Croix (eds.), Japan’s New Economy: Continuity and Change in the 21st Century. Oxford: Oxford University Press.
163 Mason, Andrew, Naohiro Ogawa, and Takehiro Fukui. 1992. Household Projections for Japan, 1985-2025: A Transition Model of Headship Rates. Tokyo: Japan Statistical Association.
Mills, Anne, Sara Bennett, Porntep Siriwanrangsun, and Viroj Tangcharoensathien. 2000. “The Response of Providers to Capitation Payment: A Case-study from Thailand.” Health Policy 51(3): 163-80.
Mitchell, Olivia S. 1990. “Aging, Job Satisfaction and Job Performance.” In I. Bluestone, R. Montgomery, and J. Owen (eds.), An Aging Workforce, pp. 242-72. Detroit: Wayne State University Press.
Miyajima, Hiroshi. 1994. “The Family Structure in Contemporary Japan.” Japanese Economic Studies 21(6) (Winter): 25-54.
Mongkolsmai, Dow. 1997. “Private Health Sector Growth and Social Security Insurance in Newbrander (ed.), Private Health Sector Growth in Asia: Issues and Implications, Chapter 6, pp. 83-107. New York: Wiley.
Murthy, N. R. Vasudeva, and Victor Ukpolo. 1994. “Aggregate Health Care Expenditures in the United States: Evidence from Cointegration Tests.” Applied Economics 26(8) (Aug.): 797-802.
Nanadakumar, A. K., Maha El-Adawy, and Marc A. Cohen. 1998. Perceptions of Health Status and Limitations in Activities among the Egyptian Elderly. Data for Decision Making Project, Harvard School of Public Health.
Nanadakumar, A. K., Mukesh Chawla, and Maryam Khan. 2000. “Utilization of Outpatient Care in Egypt and its Implication for the Role of Government in Health Care Provision.” World Development 28(1): 187-96.
Nanadakumar, A. K., Michael R. Reich, Mukesh Chawla, Perter Berman, and Winnie Yip. 2000. “Health Reform for Children: The Egyptian Experience with School Health Insurance.” Health Policy 50: 155-70.
Natividad, Josefina N., and Grace T. Cruz. 1997. “Patterns in Living Arrangements and Familiar Support for the Elderly in the Philippines.” Asia-Pacific Population Journal 12(4): 17-34.
Neumark, David, and Wendy A. Stock. 1999. “Age Discrimination Laws and Labor Market Journal of Political Economy 107(5): 1081-1125.
Newhouse, Joseph P. 1977. “Medical-care Expenditure: A Cross-national Survey.” Journal of Human Resources 12(1) (Winter): 115-25.
Newhouse, Joseph P. 1996. “Reimbursing Health Plans and Health Providers: Efficiency in Journal of Economic Literature 34 (Sept.): 1236-63.
164 O’Connell, Joan M. 1996. “The Relationship between Health Expenditures and the Age Structure of the Population in OECD Countries.” Health Economics 5(6): 573-78.
OECD. 1999. OECD Health Data 99: A Comparative Analysis of 29 Countries. Paris: OECD.
Ogawa, Naohiro, and Robert D. Retherford. 1993. “Care of the Elderly in Japan: Changing Norms and Expectations.” Journal of Marriage and the Family 55(3) (August): 585-97.
Ogawa, Naohiro, and Robert D. Retherford. 1997. “Shifting Costs of Caring for the Elderly Back Population and Development Review 23(1) (March): 59-94.
Oliver, Adam James, Naoki Ikegami, and Shunya Ikeda. 1997. “Japan’s Aging Population: Implications for Health Care.” Pharmacoeconomics 11(4) (Apr.): 306-18.
Ortega-Osona, Jose Antonio, and David Reher. 1997. “Short-term Economic Fluctuations and Demographic Behaviour: Some Examples from Twenties-Century South America.” In Georges Tapinos et al. (eds.), Demographic Responses to Economic Adjustment in Latin America. Oxford: Clarendon Press.
Palloni, Alberto, and Kenneth Hill. 1997. “The Effects of Economic Changes on Mortality by Age and Cause: Latin America, 1950-90.” In Georges Tapinos et al. (eds.), Demographic Responses to Economic Adjustment in Latin America. Oxford: Clarendon Press.
Partnerships for Health Reform. 1999. A New Egyptian Health Care Model for the 21st Century. Bethesda, MD: Abt Associates, Inc. and Ministry of Health and Population, Arab Republic of Egypt.
Pauly, Mark V. 1968. “The Economics of Moral Hazard: Comment.” American Economic Review 58(3)(1): 531-37.
Pauly, Mark V. 1974. “Overinsurance and the Public Provision of Insurance: The Roles of Moral Hazard and Adverse Selection.” Quarterly Journal of Economics 88(1): 44-62.
Pauly, Mark V. 1986. “Taxation, Health Insurance and Market Failure in the Medical Economy.” Journal of Economic Literature 24(2): 629-75.
Pingali, Prabhu L., and Hans P. Binswanger. 1987. “Intensification: A Study of the Evolution of Technoligies in Tropical Agriculture.” In D. Gale Johnson and Ronald D. Lee (eds.), Population Growth and Economic Development: Issues and Evidence, pp. 27-56. Madison, WI: University of Wisconsin Press.
Rannan-Eliya, Ravi P., Khaled H. Nada, Abeer M. Kamal, and Ahmed Ibrahim Ali. 1997. Egypt National Health Accounts 1994/95. Department of Planning, Ministry of Health and Population, Arab Republic of Egypt and Data for Decision Making Project, Harvard School of Public Health, Harvard University.
165 Robinson, James A., and T. N. Srinivasan. 1997. “Long-term Consequences of Population Growth: Technological Change, Natural Resources, and the Environment.” In Mark R. Rosenzweig and Oded Start, Handbook of Population and Family Economics. Handbook in Economics 14. Amsterdam: Elsever Science B.V., pp. 1177-1298.
Rofman, Rafael P., 2000. “Moving Social Security towards Fully Funded Schemes: Who Pays the Cost?” In Andrew Mason and Georges Tapinos (eds.), Sharing the Wealth, pp. 140-61. Oxford: The Oxford University Press.
Rothschild, Michael, and Joseph Stiglitz. 1976. “Equilibrium in Competitive Insurance Markets: An Essay in the Economics of Imperfect Information.” Quarterly Journal of Economics (90): 629-50.
Russo, Gerard. 1992. Household Expenditures for Health Care in Thailand: Estimates from the 1988 Socioeconomic Survey. Report prepared for the Asian Development Bank, Research Program on Priority Health and Population Issues R.E.T.A. 5421. Honolulu, Hawaii: East-West Center.
Russo, Gerard, and Alejandro Herrin. 1993. The Determinants of Household Health Care Expenditures in the Philippines. East-West Center Working Paper, Population Series No. 68. (Revised). Honolulu, Hawaii: East-West Center.
Russo, Gerard, Naohiro Ogawa, Wandaningsih, and Nina Sarjunani. 1995. Household expenditures for health care in Indonesia: estimates from the 1987 socioeconomic survey (SUSENAS). Report prepared for the HOMES Indonesia Project, T.A. 1140 Demographic- Economic Forecasting Study for Development Planning, Indonesia, Asian Development Bank. East-West Center.
Rust, John. 1989. “A Dynamic Programming Model of Retirement Behavior.” In David A. Wise (ed.), Studies in the Economics of Aging, pp. 359-98, NBER Project Report. Chicago: The University of Chicago Press.
Saez, Marc, and Carles Murillo. 1994. “Shared ‘Features’ in Prices: Income and Price Elasticities for Health Care Expenditures.” Health Economics 3(4): 267-79.
Samad, Abdus, and Samad Abedin. 2000. “The Implications of Asia’s Population Future for the Family and the Elderly: Bangladesh.” The Family and Older Persons in Bangladesh, Pakistan and Sri Lanka, Economic and Social Commission for Asia and the Pacific (ESCAP) Asian Population Studies Series No. 151. New York: United Nations.
Sanderson, Warren C. 1999. “Knowledge Can Improve Forecasts: A Review of Selected Socioeconomic Population Projection Models.” In Wolfgang Lutz, James W. Vaupel, and Dennis A. Ahlburg (eds.), Frontiers of Population Forecasting, a supplement to Population and Development Review (24): 88-117.
166 Saphir, Ann. 2000. “Seniors in Japan Access New Services.” Modern Healthcare (Apr. 24): 32- 33.
Shin, Youngsoo. 1998. “Health Care Systems in Transition II. Korea, Part I. An overview of Journal of Public Health Medicine 20(1): 41-46.
Simon, Julian L. 1981. The Ultimate Resource. Princeton, NJ: Princeton University Press.
Social Security Administration. 1999. Social Security Programs Throughout the World. Washington D.C.: SSA.
Solow, Robert. 1956. “A Contribution to the Theory of Economic Growth.” Quarterly Journal of Economics 70(1): 65–94.
Summers, Paul, and Robert Heston. 1991. “The Penn World Table (Mark 5): An Expanded Data Set of International Comparisons, 1950–1988.” Quarterly Journal of Economics 106(2): 145–78.
Summers, Paul, and Robert Heston. 1998. World Population Prospects: The 1998 Revision. New York: United Nations.
Supakankunti, Siripen. 2000. “Future Prospects for Voluntary Health Insurance in Thailand.” Health Policy and Planning 15(1): 85-94.
Tangcharoensathien, Viroj, Adit Laixuthai, Jitpranee Vasavit, Nuan-Anan Tantigate, Wiphant Prajuabmoh-Ruffolo, Duangkamol Vimolkit, and Jongkol Lertiendumrong. 1999. “National Health Accounts Development: Lessons from Thailand.” Health Policy and Planning 14(4): 342- 53.
Tangcharoensathien, Viroj, Anuwat Supachutikul, Jongkol Lertiendumrong. 1999. “The Social Security Scheme in Thailand: What Lessons Can Be Drawn?” Social Science & Medicine 48(7): 913-23.
Thailand, National Economic and Social Development Board (NESDB), Human Resources Planning Division. 1995. Population Projections for Thailand, 1990-2020. Bangkok: NESDB.
Thailand, The NESDB Working Group on HIV/AIDS Projection. Projections for HIV/AIDS in Thailand: 1987-2020. Bangkok: Thai Red Cross Society Program on AIDS.
Toh, Mun-Heng. 1997. “Saving, Capital Formation, and Economic Growth in Singapore.” East- West Center Working Papers, Population Series No. 88-8. Honolulu: East-West Center.
Turner, John. 1997. “Retirement Income Systems for Different Economic, Demographic and Political Environments.” Tripartite regional consultation with Asian countries, Bangkok, October 1997, ILO Social Security Department
167 U.S. Bureau of the Census. 1999. World Population Profile: 1998, Report WP/98. Washington, D.C.: U.S. Government Printing Office.
U.S. Department of Justice, Immigration and Naturalization Service. 1997 and 1998. Statistical Yearbook of the Immigration and Naturalization Service.
United Nations, various years. UNAIDS/WHO Epidemiological Fact Sheets (URL: www.unaids.org).
United Nations. 1998. World Population Monitoring 1997. New York: United Nations.
United Nations. 1998. World Population Prospects: The 1998 Revision. New York: United Nations.
United Nations. 1999. Population Ageing (wall chart). New York: United Nations.
Valdés-Prieto, Salvador, 2000. “The Political Economy of Two Chilean Pension Systems.” In Andrew Mason and Georges Tapinos (eds.), Sharing the Wealth, pp. 162-97. Oxford: The Oxford University Press. van Doorslaer, Eddy, and Adam Wagstaff. 1992. “Equity in the Delivery of Health Care: Some International Comparisons.” Journal of Health Economics 11(4) (Dec.): 389-411. van Doorslaer, Eddy, Adam Wagstaff, Hattem van der Burg, et al. 1999. “The Redistributive Effect of Health Care Finance in Twelve OECD Countries.” Journal of Health Economics 18(3) (June): 2291-313.
Vogel, Ronald J. 1980. “The Tax Treatment of Health Insurance Premiums as a Cause of Pauly (ed.), National Health Insurance: What Now, What Later, What Never?, pp. 220-49. Washington, D.C.: American Enterprise Institute.
Wagstaff, Adam, and Eddy van Doorslaer. 1992. “Equity in the Finance of Health Care: Some International Comparisons.” Journal of Health Economics 11(4) (Dec.): 361-87.
Wagstaff, Adam, Eddy van Doorslaer, Hattem van der Burg, et al. 1999. “Equity in the Finance of Health Care: Some Further International Comparisons.” Journal of Health Economics 18(3) (June): 263-90.
Watts, Jonathon. 2000. “Chaos Greets Birth of Insurance System for Japan’s Elderly.” The Lancet 355 (Apr. 8): 1249.
Weinstein, W., T. H. Sun, M. C. Chang, and R. Freedman. 1994. “Co-Residence and Other Ties Linking Couples and their Parents.” In Arland Thornton and Hui-Li Sheng Lin (eds.), Social Change & the Family in Taiwan. Chicago: The University of Chicago Press.
168 Williamson, Jeffrey, and Matthew Higgins. 1997. “The Accumulation and Demography Conference on Population and the Asian Economic Miracle. East- West Center, Honolulu, Hawaii, January 7-10. In press.
World Bank. 1994. Averting the Old Age Crisis. Washington, D.C.: World Bank.
World Bank. 1999. 1999 World Development Indicators, CD ROM. Washington D.C.: World Bank.
World Bank. World Bank Database.
World Health Organization. 1998. World Health Report 1998: Life in the 21st Century –A Vision for All. Geneva: WHO.
World Health Organization. 1999. World Health Report 1999: Making a Difference. Geneva: WHO.
World Health Organization. 2000. World Health Report 2000: Health Systems— Improving Performance. Geneva: WHO.
Yang, Bong-Min. 1996. “The Role of Health Insurance in the Growth of the Private Health International Journal of Health Planning and Management 11(3): 231-52.
Yap Mui Ten. 1997. “Population Policies and Programs in Singapore.” East-West Center Working Papers, Population Series, No. 88-22, Honolulu, August.
Zeckhauser, Richard. 1970. “Medical Insurance: A Case Study of the Tradeoff between Risk Spreading and Appropriate Incentives.” Journal of Economic Theory 2(1): 10-16.
Zlotnik, Hania. 1999. World Population Prospects: The 1998 Revision, Joint ECE-Eurostat Work Session on Demographic Projections, Working Paper No. 19, ISTAT, Perugia, Italy, 3-7 May.
Zweifel, Peter, Stefan Felder, and Markus Meiers. 1999. “Ageing Population and Health Care Expenditure: A Red Herring?” Health Economics 8(6): 485-96.
169 Appendix A. Population projections – technical notes
The United Nations Population Division published updated population projections in 1999.
Three variants are provided that make different assumptions about fertility trends as discussed in the text. The UN projections do not consider the implications of different mortality trends.
Alternative scenarios were prepared so as to assess the impact of changes in the pace of mortality change.
For each of the three UN variants (high, medium and low fertility), two new projections were constructed – a high and a low mortality variant. The high mortality variant assumes that projected increases in life expectancy occur half as rapidly as in the UN projections. The low mortality variant assumes that increases occur twice as rapidly as in the UN projections. In both cases, changes in the age structure of survival follow the same pattern as in the UN projections.
The chief advantages of this approach are its simplicity, the variants could be constructed using information that is publicly available, and its consistency with the UN methodology. There are several disadvantages. More complex mortality variants are not considered. For example, the implications of more rapid improvements in survival chances at older ages and slower improvements at younger ages are not considered. Also, for the low mortality variant we can construct projections only to 2025.
The alternative variants are constructed using a simple device. Survival rates were calculated from the population and birth data that are available in the UN variants. All variants assume that the survival rates for 1995-2000 are the same. Increases in survival rates that occur during a five-year period in the UN variants, require ten years in the high mortality variant.
Increases in survival rates that occur over a ten-year period in the UN variants require only five years in the low mortality variants. Hence, life expectancy in the high mortality variant in years
170 2015-20 is the same as life expectancy in 2005-10 in the standard mortality variant. Life expectancy in the low mortality variant in years 2020-25 is the same as life expectancy in the
2045-50 in the standard mortality variant. For the high mortality variant, survival rates are linearly interpolated as is necessary.
Varying survival rates also influences the number of births because of changes in the number of women in the childbearing ages. The number of births was adjusted so as to maintain the ratio of the number of births to the population aged 25-39 across alternative mortality variants. A summary of the alternative projections is provided below for the regions of Asia and for seven ANE countries. More detailed results are available.
171 Appendix B. Economically Active Population Estimates and Projections
Historical estimates and projections of the economically active population rely on the
ILO, 1996 Economically Active Population Estimates and Projections (EAPEP). The EAPEP provides estimates and projections of labor force participation rates (activity rates) by sex and five-year age group for the period 1950-2010 at ten-year intervals and for 1995.
The economically active population comprises all men and women who supply labor for the production of economic goods and services as defined by the UN systems of national accounts and balances, during a specified time period. Activity rates for each age group are the ratio, expressed as a percentage, of the economically active population in a given age group to the population in that age group.
Data on labor force are drawn from population censuses and especially from sample surveys of the economically active population. The labor force data are adjusted where necessary, to conform to a standard concept of economically active population, which consists of all employed and unemployed persons. Historical estimates and projections of the labor force are obtained by applying ILO estimates of activity rates to the population estimates and projections prepared by the UN Population Division. The most recent ILO results rely on the
UN’s 1996 population estimates and projections. The labor force results presented here are updated using activity rates from the ILO and more recent population data, the 1998 Revision.
ILO projections are based on projected activity rates for men and women in five-year age groups. The youngest age group is 10-14 and the oldest age group is 65 and older. Activity rates were projected for 1995, 2000, and 2010 by extrapolating trends in the historical data. A variety
172 of functional forms were used. However, the trends were constrained by assumptions about the likely course of activity rates. The assumptions are summarized by ILO as follows (ILO, 1996):
Men: Activity rates for all age groups continue to decrease, more markedly in the
15-19 age group and 55 years and over. However, in the more developed
countries, which registered sharp drops in activity rates in both age groups
concerned over the last two decades, the rate of decrease should gradually slow
down.
Women: Activity rates increase in the great majority of countries or territories
concerned in the 25-54 age group, irrespective of the trend registered over the
period as a whole or in the last decade. The assumption also predicts that the
profile of women’s activity rates will move closer to that of men, although female
levels will not exceed male levels. For a very small number of countries and/or
territories in Asia, characterized by very high female activity rates, the decline
registered between 1950 and 1990 continues but a much slower rate of decrease.
Two problems were addressed in employing the EAPEP projections of activity rates in this work. The first is that the EAPEP does not provide estimates and projections of activity rates for the period 2020-2050. Because the assumptions made by ILO were designed to work only until 2010, it is unlikely that the projections of activity rates for the period 2020-2050 would follow the same patterns as those for the period 1950-2010. One obvious difficulty of projecting beyond 2010 is that the rate of decrease/increase must slow as they reach low/high
173 levels, but these natural constraints are not incorporated into the ILO methodology. Of course, there are many other possibilities that are not considered. To prepare long-term projections, we made a strong, but simple assumption, that age- and sex-specific activity rates remain constant after 2010. Thus, changes in projected labor force aggregates after 2010 are due entirely to changes in the age- and sex-distributions of the populations, not due to changes in activity rates.
The second problem is the use of 65 and older as the upper age group by the ILO. This is problem for countries experiencing substantial changes in the age-distributions of their elderly populations because activity rates drop rapidly with age. As a simple expedient, we assume that labor force participation rates of the elderly decrease linearly beginning at age 65 and reaches zero at age 90. Thus, changes in the age distribution of the 65 and older population influence the average activity rate of that population. Given the crude approximation of this approach, however, we report labor force values only for the 65 and older group and not for sub-groups of the elderly population.
Our projections for the period 2020-2050 will overestimate or underestimate the actual labor force participation to the extent that the trends in activity rates between 1950 and the early
1990s persist into the future. It seems likely that the errors for prime age males will be relatively small. The young and the elderly, however, have registered substantial drops in their activity rates that may persist into the future. For the elderly, much will depend on employment opportunities, improvements in standards of living, health conditions, and the evolution of public and private pension programs.
Young females have also experienced declining activity rates that may drop below levels projected here. Activity rates among women in the prime working ages have increased. If this continues, our long-range projections will understate the female labor force in these ages.
174 Activity rates for older women will be governed by the same forces that apply to men. In addition, as women more actively participate in the labor force at younger ages, their labor force behavior at older ages may become similar to that of the behavior of men. Hence, the trend in participation by older women is very much an open question.
175 Appendix C. Methodology for forecasting the proportion widowed.
Forecasts of the proportion widowed are based on a model first applied to Japan (Mason, Ogawa,
s and Fukui, 1992). The variable analyzed is the transition rate, Dwat , where
s s s Dwat = wa+5,t +5 - wat
The change in the proportion currently widowed during a five-year span for a cohort of men or women (s), age a, in year t. If mortality rates were independent of marital status and neglecting remarriage, the change in the proportion widowed could be calculated directly knowing only the proportion married at the beginning of a period and the proportion experiencing the death of a spouse during any subsequent time period. The latter could be calculated from the joint age distribution of husbands and wives at the beginning of the period and age- and sex-specific mortality rates. The change in the proportion widowed differs from the directly calculated value for several reasons. First, mortality rates may not be independent of marital status. In particular, the probability of dying may be elevated for those who have recently experienced the loss of a spouse. Under these circumstances the proportion widowed will increase by less than the proportion experiencing the death of a spouse. Second, some widows or widowers will remarry. Again, the proportion widowed will increase by less than the proportion experiencing the death of a spouse. The forecasting model used here allows for the impacts of marital-status-dependent mortality and remarriage using regression techniques. To simplify the analysis, we assume no remarriage for the moment, thus, the change in the absolute number of widows, can be represented by the following accounting identity.
1 2 Wa+5,t +5 = qatWat + qatd at M at
1 1 1 where Wat and M at are the number of widows and married women aged a in year t, qat is the
2 probability that a widow aged a in year t will survive five additional years, qat , is the survival
176 probability for a woman who is widowed during the five year period, and dat is the five year death rate for husbands of women aged a in year . Dividing both sides by the total number of women aged a in year t and representing the survival rate for women in general by qat , this yields
1 2 qat wa+5,t +5 = qat wat + qatd atmat
where qat = N a+5,t +5 / N at , w and m are the proportions widowed and married, respectively, and
1 N at is a population aged a in year t. Finally letting e at = qat - qat and rearranging terms, the transition rate is given by:
2 e at qat Dwat = wat + dat mat qat qat
The transition rate can be represented, then, as a homogeneous linear function of the proportion widowed and the product of the proportion married and the death rate for husbands. If mortality
1 2 is not systematically related to marital status, i.e., qat = qat = qat , then the expression for the transition rate simplifies to Dwat = d atmat . A similar expression is used to model the transition rate for men. The death rates for husbands and wives are not available but can be approximated using information about the mortality rates for men and women and the age distribution of husbands
s and wives. Death rates for men and women, d at , are approximated using intercensal survival techniques, i.e., the death rate for persons aged a and sex s between t and t+5 is calculated as: N s - N s d s at a+5,t +5 at = s N at
The joint age distributions of husbands and wives are based on tabulations of the joint distributions of husbands and wives heading intact households at the beginning of each
177 intercensal period. Letting haxt represent the proportion of married men aged x who have wives aged a, then the death rate for wives of men aged x is given by :
v+ ˆ m f d xt = åhaxtd at a=u-u+5 where u~u+5 is the youngest intercensal age group applicable to the model, and v+ is the oldest (open ended) age group. The death rate of husbands can be calculated in a symmetrical way. Estimation of the transition equation requires the imposition of restrictions on the age pattern of marital status related mortality differences. In particular, we will assume that the relative differences in survival between marital status groups do not vary with age. In the case of the mortality of spouses, we assume that a = e at / qa1 and anticipate that a will be greater than ˆ zero, and d at = gdat while anticipating that g will be greater than one. And in the case of survival of those who become widowed during the period and survival in general, we assume
2 that b = qat / qat and anticipate a value substantially less than one but greater than zero. Our final estimation is arrived at by substituting into the transition rate equation mentioned before yielding :
s s ˆ s s Dwat = awat + bgd atmat
The equation is estimated separately for men and women, in each country for the periods for which data are available. (For Thailand, due to data availability, the model is estimated for 10-year rather than 5-year periods. The model is estimated using ordinary least squares regression. The model does not yield transition rates for the uppermost age category. Consequently, we rely on a crude approximation, assuming the transition rate to the oldest age interval is a constant proportion of the transition rate to the oldest closed interval. The proportions used are the average values for the period analyzed. For example, in most instances the upper age interval is 85+ in which case:
178 s s Dw80~84,t = aˆw75~79,t
The estimated parameters for the four countries for which forecasts were prepared are:
a bg aˆ
Indonesia Males 0.469 0.000 1.140 Female 0.123 0.485 0.188
South Korea Males 0.091 0.418 0.457 Female 0.030 0.741 0.056
Philippines Males 0.422 0.000 0.820 Female 0.126 0.486 0.328
Thailand Males 1.009 0.179 2.449 Female 0.411 0.672 0.948
The forecast uses intercensal survival rates calculated from UN population projections and the most recently available data on the joint age-distributions of household heads and their wives.
179 Table A1
Summary of Alternative Population Projections Asia Medium Fertility Scenario
Population (1000s) Percentage Year 0-14 15-64 65+ 0-14 15-64 65+ Standard Mortality Scenario 1995 1,029,982 2,061,447 176,471 31.5 63.1 5.4 2000 1,036,240 2,250,946 207,301 29.7 64.4 5.9 2005 1,008,070 2,459,522 238,532 27.2 66.4 6.4 2010 990,796 2,644,998 269,518 25.4 67.7 6.9 2015 978,901 2,800,305 315,118 23.9 68.4 7.7 2020 972,384 2,914,904 383,936 22.8 68.2 9.0 2025 963,176 3,010,679 454,718 21.7 68.0 10.3 2030 947,630 3,073,198 542,084 20.8 67.3 11.9 2035 931,674 3,099,758 645,479 19.9 66.3 13.8 2040 922,368 3,109,737 738,446 19.3 65.2 15.5 2045 918,890 3,123,276 800,035 19.0 64.5 16.5 2050 913,671 3,113,548 863,601 18.7 63.6 17.7 High Mortality Scenario 1995 1,029,982 2,061,447 176,471 31.5 63.1 5.4 2000 1,036,240 2,250,946 207,301 29.7 64.4 5.9 2005 1,006,207 2,457,291 236,949 27.2 66.4 6.4 2010 984,275 2,638,992 264,688 25.3 67.9 6.8 2015 967,863 2,788,150 305,139 23.8 68.7 7.5 2020 958,376 2,893,151 366,691 22.7 68.6 8.7 2025 946,337 2,975,593 428,612 21.8 68.4 9.9 2030 927,625 3,024,796 504,424 20.8 67.9 11.3 2035 907,431 3,039,731 592,789 20.0 67.0 13.1 2040 893,609 3,038,353 669,413 19.4 66.0 14.5 2045 885,971 3,039,565 716,682 19.1 65.5 15.4 2050 877,942 3,018,109 766,636 18.8 64.7 16.4 Low Mortality Scenario 1995 1,029,982 2,061,447 176,471 31.5 63.1 5.4 2000 1,036,240 2,250,946 207,301 29.7 64.4 5.9 2005 1,012,712 2,463,057 241,661 27.2 66.3 6.5 2010 1,000,197 2,656,072 278,752 25.4 67.5 7.1 2015 993,654 2,822,603 332,764 23.9 68.0 8.0 2020 991,221 2,949,849 412,879 22.8 67.8 9.5 2025 984,334 3,057,071 496,731 21.7 67.4 10.9 Table A2
Summary of Alternative Population Projections Asia High Fertility Scenario
Population (1000s) Percentage Year 0-14 15-64 65+ 0-14 15-64 65+ Standard Mortality Scenario 1995 1,029,982 2,061,447 176,471 31.5 63.1 5.4 2000 1,049,734 2,250,946 207,301 29.9 64.2 5.9 2005 1,052,794 2,459,522 238,532 28.1 65.6 6.4 2010 1,081,465 2,644,998 269,518 27.1 66.2 6.7 2015 1,115,274 2,813,492 315,118 26.3 66.3 7.4 2020 1,149,251 2,958,890 383,936 25.6 65.9 8.5 2025 1,180,011 3,100,155 454,718 24.9 65.5 9.6 2030 1,207,120 3,221,076 542,084 24.3 64.8 10.9 2035 1,237,872 3,318,450 645,479 23.8 63.8 12.4 2040 1,277,252 3,413,288 738,446 23.5 62.9 13.6 2045 1,323,067 3,527,065 800,035 23.4 62.4 14.2 2050 1,366,965 3,633,705 863,601 23.3 62.0 14.7 High Mortality Scenario 1995 1,029,982 2,061,447 176,471 31.5 63.1 5.4 2000 1,049,734 2,250,946 207,301 29.9 64.2 5.9 2005 1,050,759 2,457,291 236,949 28.1 65.6 6.3 2010 1,074,218 2,638,992 264,688 27.0 66.3 6.7 2015 1,102,611 2,801,240 305,139 26.2 66.6 7.2 2020 1,132,642 2,936,587 366,691 25.5 66.2 8.3 2025 1,159,229 3,063,612 428,612 24.9 65.9 9.2 2030 1,181,302 3,169,780 504,424 24.3 65.3 10.4 2035 1,205,130 3,253,433 592,789 23.9 64.4 11.7 2040 1,236,764 3,333,905 669,413 23.6 63.6 12.8 2045 1,274,961 3,431,202 716,682 23.5 63.3 13.2 2050 1,312,711 3,520,447 766,636 23.4 62.9 13.7 Low Mortality Scenario 1995 1,029,982 2,061,447 176,471 31.5 63.1 5.4 2000 1,049,734 2,250,946 207,301 29.9 64.2 5.9 2005 1,057,747 2,463,057 241,661 28.1 65.5 6.4 2010 1,091,902 2,656,127 278,752 27.1 66.0 6.9 2015 1,132,217 2,836,109 332,764 26.3 65.9 7.7 2020 1,171,722 2,994,919 412,879 25.6 65.4 9.0 2025 1,206,194 3,149,052 496,731 24.9 64.9 10.2 Table A3
Summary of Alternative Population Projections Asia Low Fertility Scenario
Population (1000s) Percentage Year 0-14 15-64 65+ 0-14 15-64 65+ Standard Mortality Scenario 1995 1,029,982 2,061,447 176,471 31.5 63.1 5.4 2000 1,022,125 2,250,946 207,301 29.4 64.7 6.0 2005 958,581 2,459,522 238,532 26.2 67.3 6.5 2010 886,467 2,644,998 269,518 23.3 69.6 7.1 2015 820,698 2,786,541 315,118 20.9 71.0 8.0 2020 775,793 2,866,209 383,936 19.3 71.2 9.5 2025 739,401 2,907,647 454,718 18.0 70.9 11.1 2030 698,028 2,903,053 542,084 16.8 70.1 13.1 2035 649,758 2,856,763 645,479 15.6 68.8 15.5 2040 606,277 2,785,794 738,446 14.7 67.4 17.9 2045 572,770 2,707,176 800,035 14.0 66.3 19.6 2050 543,577 2,593,419 863,601 13.6 64.8 21.6 High Mortality Scenario 1995 1,029,982 2,061,447 176,471 31.5 63.1 5.4 2000 1,022,125 2,250,946 207,301 29.4 64.7 6.0 2005 956,961 2,457,291 236,949 26.2 67.3 6.5 2010 880,955 2,638,992 264,688 23.3 69.7 7.0 2015 811,817 2,774,468 305,139 20.9 71.3 7.8 2020 764,804 2,845,172 366,691 19.2 71.5 9.2 2025 726,493 2,874,510 428,612 18.0 71.3 10.6 2030 683,303 2,858,345 504,424 16.9 70.6 12.5 2035 633,098 2,802,592 592,789 15.7 69.6 14.7 2040 587,935 2,723,145 669,413 14.8 68.4 16.8 2045 552,973 2,636,117 716,682 14.2 67.5 18.3 2050 523,073 2,515,941 766,636 13.7 66.1 20.1 Low Mortality Scenario 1995 1,029,982 2,061,447 176,471 31.5 63.1 5.4 2000 1,022,125 2,250,946 207,301 29.4 64.7 6.0 2005 962,745 2,463,057 241,661 26.3 67.2 6.6 2010 894,541 2,656,005 278,752 23.4 69.4 7.3 2015 832,997 2,808,294 332,764 21.0 70.7 8.4 2020 790,800 2,899,462 412,879 19.3 70.7 10.1 2025 755,318 2,950,568 496,731 18.0 70.2 11.8 Table A4
Summary of Aging, Alternative Mortality Scenarios Asia
Percentage 65 and older Old Age Dependency Ratio Year Medium High Low Medium High Low Medium Fertility Scenario 1995 5.4 5.4 5.4 0.086 0.086 0.086 2000 5.9 5.9 5.9 0.092 0.092 0.092 2005 6.4 6.4 6.5 0.097 0.096 0.098 2010 6.9 6.8 7.1 0.102 0.100 0.105 2015 7.7 7.5 8.0 0.113 0.109 0.118 2020 9.0 8.7 9.5 0.132 0.127 0.140 2025 10.3 9.9 10.9 0.151 0.144 0.162 2030 11.9 11.3 0.176 0.167 2035 13.8 13.1 0.208 0.195 2040 15.5 14.5 0.237 0.220 2045 16.5 15.4 0.256 0.236 2050 17.7 16.4 0.277 0.254 High Fertility Scenario 1995 5.4 5.4 5.4 0.086 0.086 0.086 2000 5.9 5.9 5.9 0.092 0.092 0.092 2005 6.4 6.3 6.4 0.097 0.096 0.098 2010 6.7 6.7 6.9 0.102 0.100 0.105 2015 7.4 7.2 7.7 0.112 0.109 0.117 2020 8.5 8.3 9.0 0.130 0.125 0.138 2025 9.6 9.2 10.2 0.147 0.140 0.158 2030 10.9 10.4 0.168 0.159 2035 12.4 11.7 0.195 0.182 2040 13.6 12.8 0.216 0.201 2045 14.2 13.2 0.227 0.209 2050 14.7 13.7 0.238 0.218 Low Fertility Scenario 1995 5.4 5.4 5.4 0.086 0.086 0.086 2000 6.0 6.0 6.0 0.092 0.092 0.092 2005 6.5 6.5 6.6 0.097 0.096 0.098 2010 7.1 7.0 7.3 0.102 0.100 0.105 2015 8.0 7.8 8.4 0.113 0.110 0.118 2020 9.5 9.2 10.1 0.134 0.129 0.142 2025 11.1 10.6 11.8 0.156 0.149 0.168 2030 13.1 12.5 0.187 0.176 2035 15.5 14.7 0.226 0.212 2040 17.9 16.8 0.265 0.246 2045 19.6 18.3 0.296 0.272 2050 21.6 20.1 0.333 0.305 Table A5
Total Dependency Ratio Asia
Fertility: High Med Low High Med Low High Med Low Mortality: High High High Med Med Med Low Low Low
1995 0.585 0.585 0.585 0.585 0.585 0.585 0.585 0.585 0.585 2000 0.558 0.552 0.546 0.558 0.552 0.546 0.558 0.552 0.546 2005 0.524 0.506 0.486 0.525 0.507 0.487 0.528 0.509 0.489 2010 0.507 0.473 0.434 0.511 0.476 0.437 0.516 0.482 0.442 2015 0.503 0.457 0.403 0.508 0.462 0.408 0.519 0.470 0.415 2020 0.511 0.458 0.398 0.518 0.465 0.405 0.537 0.476 0.415 2025 0.518 0.462 0.402 0.527 0.471 0.411 0.557 0.484 0.424 2030 0.532 0.473 0.416 0.543 0.485 0.427 2035 0.553 0.494 0.437 0.568 0.509 0.453 2040 0.572 0.514 0.462 0.591 0.534 0.483 2045 0.580 0.527 0.482 0.602 0.550 0.507 2050 0.591 0.545 0.513 0.614 0.571 0.543 Table A6
Summary of Alternative Population Projections East Asia Medium Fertility Scenario
Population (1000s) Percentage Year 0-14 15-64 65+ 0-14 15-64 65+ Standard Mortality Scenario 1995 361,439 963,686 97,140 25.4 67.8 6.8 2000 354,957 1,015,871 114,366 23.9 68.4 7.7 2005 327,599 1,080,483 130,440 21.3 70.2 8.5 2010 316,241 1,125,580 145,916 19.9 70.9 9.2 2015 312,869 1,149,265 171,496 19.2 70.3 10.5 2020 311,515 1,148,553 210,223 18.6 68.8 12.6 2025 304,358 1,149,867 241,040 18.0 67.8 14.2 2030 291,858 1,134,311 282,797 17.1 66.4 16.5 2035 282,187 1,095,974 335,677 16.5 63.9 19.6 2040 277,990 1,057,304 375,746 16.2 61.8 22.0 2045 275,472 1,037,877 384,743 16.2 61.1 22.7 2050 270,398 1,016,758 388,315 16.1 60.7 23.2 High Mortality Scenario 1995 361,439 963,686 97,140 25.4 67.8 6.8 2000 354,957 1,015,871 114,366 23.9 68.4 7.7 2005 327,276 1,079,904 129,540 21.3 70.3 8.4 2010 315,396 1,123,905 143,386 19.9 71.0 9.1 2015 311,321 1,146,306 166,430 19.2 70.6 10.2 2020 309,381 1,143,817 201,897 18.7 69.1 12.2 2025 301,682 1,142,683 228,674 18.0 68.3 13.7 2030 288,753 1,124,791 265,542 17.2 67.0 15.8 2035 278,599 1,084,833 311,962 16.6 64.8 18.6 2040 273,788 1,044,896 344,518 16.5 62.8 20.7 2045 270,675 1,023,939 347,610 16.5 62.4 21.2 2050 265,181 1,001,392 346,556 16.4 62.1 21.5 Low Mortality Scenario 1995 361,439 963,686 97,140 25.4 67.8 6.8 2000 354,957 1,015,871 114,366 23.9 68.4 7.7 2005 328,119 1,081,576 132,051 21.3 70.2 8.6 2010 317,660 1,128,340 150,516 19.9 70.7 9.4 2015 315,388 1,154,531 179,999 19.1 70.0 10.9 2020 314,760 1,156,364 224,474 18.6 68.2 13.2 2025 307,728 1,159,871 261,456 17.8 67.1 15.1 Table A7
Summary of Alternative Population Projections East Asia High Fertility Scenario
Population (1000s) Percentage Year 0-14 15-64 65+ 0-14 15-64 65+ Standard Mortality Scenario 1995 361,439 963,686 97,140 25.4 67.8 6.8 2000 356,805 1,015,871 114,366 24.0 68.3 7.7 2005 336,329 1,080,483 130,440 21.7 69.8 8.4 2010 334,398 1,125,580 145,916 20.8 70.1 9.1 2015 343,110 1,151,099 171,496 20.6 69.1 10.3 2020 353,688 1,157,226 210,223 20.5 67.2 12.2 2025 359,200 1,167,923 241,040 20.3 66.0 13.6 2030 357,581 1,166,223 282,797 19.8 64.5 15.7 2035 357,635 1,146,578 335,677 19.4 62.3 18.2 2040 364,025 1,129,866 375,746 19.5 60.4 20.1 2045 373,381 1,135,048 384,743 19.7 59.9 20.3 2050 379,632 1,142,162 388,315 19.9 59.8 20.3 High Mortality Scenario 1995 361,439 963,686 97,140 25.4 67.8 6.8 2000 356,805 1,015,871 114,366 24.0 68.3 7.7 2005 335,986 1,079,904 129,540 21.7 69.9 8.4 2010 333,489 1,123,905 143,386 20.8 70.2 9.0 2015 341,407 1,148,137 166,430 20.6 69.3 10.1 2020 351,270 1,152,450 201,897 20.6 67.6 11.8 2025 356,052 1,160,623 228,674 20.4 66.5 13.1 2030 353,766 1,156,431 265,542 19.9 65.1 15.0 2035 353,029 1,134,908 311,962 19.6 63.1 17.3 2040 358,406 1,116,553 344,518 19.7 61.4 18.9 2045 366,726 1,119,694 347,610 20.0 61.1 19.0 2050 372,127 1,124,679 346,556 20.2 61.0 18.8 Low Mortality Scenario 1995 361,439 963,686 97,140 25.4 67.8 6.8 2000 356,805 1,015,871 114,366 24.0 68.3 7.7 2005 336,877 1,081,576 132,051 21.7 69.8 8.5 2010 335,915 1,128,347 150,516 20.8 69.9 9.3 2015 345,880 1,156,409 179,999 20.6 68.7 10.7 2020 357,405 1,165,192 224,474 20.5 66.7 12.8 2025 363,253 1,178,294 261,456 20.1 65.4 14.5 Table A8
Summary of Alternative Population Projections East Asia Low Fertility Scenario
Population (1000s) Percentage Year 0-14 15-64 65+ 0-14 15-64 65+ Standard Mortality Scenario 1995 361,439 963,686 97,140 25.4 67.8 6.8 2000 353,112 1,015,871 114,366 23.8 68.5 7.7 2005 315,287 1,080,483 130,440 20.7 70.8 8.5 2010 285,985 1,125,580 145,916 18.4 72.3 9.4 2015 262,698 1,147,434 171,496 16.6 72.5 10.8 2020 250,116 1,136,322 210,223 15.7 71.2 13.2 2025 238,327 1,119,778 241,040 14.9 70.0 15.1 2030 221,670 1,082,565 282,797 14.0 68.2 17.8 2035 203,342 1,022,690 335,677 13.0 65.5 21.5 2040 188,265 961,640 375,746 12.3 63.0 24.6 2045 177,179 916,552 384,743 12.0 62.0 26.0 2050 167,489 865,469 388,315 11.8 60.9 27.3 High Mortality Scenario 1995 361,439 963,686 97,140 25.4 67.8 6.8 2000 353,112 1,015,871 114,366 23.8 68.5 7.7 2005 314,988 1,079,904 129,540 20.7 70.8 8.5 2010 285,241 1,123,905 143,386 18.4 72.4 9.2 2015 261,404 1,144,479 166,430 16.6 72.8 10.6 2020 248,417 1,131,639 201,897 15.7 71.5 12.8 2025 236,264 1,112,782 228,674 15.0 70.5 14.5 2030 219,364 1,073,491 265,542 14.1 68.9 17.0 2035 200,833 1,012,350 311,962 13.2 66.4 20.5 2040 185,529 950,483 344,518 12.5 64.2 23.3 2045 174,233 904,445 347,610 12.2 63.4 24.4 2050 164,426 852,709 346,556 12.1 62.5 25.4 Low Mortality Scenario 1995 361,439 963,686 97,140 25.4 67.8 6.8 2000 353,112 1,015,871 114,366 23.8 68.5 7.7 2005 315,771 1,081,576 132,051 20.6 70.7 8.6 2010 287,204 1,128,329 150,516 18.3 72.0 9.6 2015 264,767 1,152,616 179,999 16.6 72.2 11.3 2020 252,660 1,143,846 224,474 15.6 70.6 13.8 2025 240,866 1,129,083 261,456 14.8 69.2 16.0 Table A9
Summary of Aging, Alternative Mortality Scenarios East Asia
Percentage 65 and older Old Age Dependency Ratio Year Medium High Low Medium High Low Medium Fertility Scenario 1995 6.8 6.8 6.8 0.101 0.101 0.101 2000 7.7 7.7 7.7 0.113 0.113 0.113 2005 8.5 8.4 8.6 0.121 0.120 0.122 2010 9.2 9.1 9.4 0.130 0.128 0.133 2015 10.5 10.2 10.9 0.149 0.145 0.156 2020 12.6 12.2 13.2 0.183 0.177 0.194 2025 14.2 13.7 15.1 0.210 0.200 0.225 2030 16.5 15.8 0.249 0.236 2035 19.6 18.6 0.306 0.288 2040 22.0 20.7 0.355 0.330 2045 22.7 21.2 0.371 0.339 2050 23.2 21.5 0.382 0.346 High Fertility Scenario 1995 6.8 6.8 6.8 0.101 0.101 0.101 2000 7.7 7.7 7.7 0.113 0.113 0.113 2005 8.4 8.4 8.5 0.121 0.120 0.122 2010 9.1 9.0 9.3 0.130 0.128 0.133 2015 10.3 10.1 10.7 0.149 0.145 0.156 2020 12.2 11.8 12.8 0.182 0.175 0.193 2025 13.6 13.1 14.5 0.206 0.197 0.222 2030 15.7 15.0 0.242 0.230 2035 18.2 17.3 0.293 0.275 2040 20.1 18.9 0.333 0.309 2045 20.3 19.0 0.339 0.310 2050 20.3 18.8 0.340 0.308 Low Fertility Scenario 1995 6.8 6.8 6.8 0.101 0.101 0.101 2000 7.7 7.7 7.7 0.113 0.113 0.113 2005 8.5 8.5 8.6 0.121 0.120 0.122 2010 9.4 9.2 9.6 0.130 0.128 0.133 2015 10.8 10.6 11.3 0.149 0.145 0.156 2020 13.2 12.8 13.8 0.185 0.178 0.196 2025 15.1 14.5 16.0 0.215 0.205 0.232 2030 17.8 17.0 0.261 0.247 2035 21.5 20.5 0.328 0.308 2040 24.6 23.3 0.391 0.362 2045 26.0 24.4 0.420 0.384 2050 27.3 25.4 0.449 0.406 Table A10
Total Dependency Ratio East Asia
Fertility: High Med Low High Med Low High Med Low Mortality: High High High Med Med Med Low Low Low
1995 0.476 0.476 0.476 0.476 0.476 0.476 0.476 0.476 0.476 2000 0.464 0.462 0.460 0.464 0.462 0.460 0.464 0.462 0.460 2005 0.431 0.423 0.412 0.432 0.424 0.413 0.434 0.425 0.414 2010 0.424 0.408 0.381 0.427 0.411 0.384 0.431 0.415 0.388 2015 0.442 0.417 0.374 0.447 0.421 0.378 0.455 0.429 0.386 2020 0.480 0.447 0.398 0.487 0.454 0.405 0.503 0.466 0.417 2025 0.504 0.464 0.418 0.514 0.474 0.428 0.539 0.491 0.445 2030 0.536 0.493 0.452 0.549 0.507 0.466 2035 0.586 0.544 0.507 0.605 0.564 0.527 2040 0.630 0.592 0.558 0.655 0.618 0.587 2045 0.638 0.604 0.577 0.668 0.636 0.613 2050 0.639 0.611 0.599 0.672 0.648 0.642 Table A11
Summary of Alternative Population Projections Southeast Asia Medium Fertility Scenario
Population (1000s) Percentage Year 0-14 15-64 65+ 0-14 15-64 65+ Standard Mortality Scenario 1995 163,217 296,477 20,768 34.0 61.7 4.3 2000 163,276 330,762 24,499 31.5 63.8 4.7 2005 160,964 364,308 28,810 29.1 65.7 5.2 2010 158,354 396,836 32,956 26.9 67.5 5.6 2015 156,616 426,776 38,004 25.2 68.7 6.1 2020 156,544 450,483 46,347 24.0 68.9 7.1 2025 157,235 468,040 58,241 23.0 68.5 8.5 2030 156,887 481,935 71,918 22.1 67.8 10.1 2035 155,859 491,305 87,409 21.2 66.9 11.9 2040 154,670 497,301 102,944 20.5 65.9 13.6 2045 154,146 500,181 117,510 20.0 64.8 15.2 2050 154,164 500,281 131,080 19.6 63.7 16.7 High Mortality Scenario 1995 163,217 296,477 20,768 34.0 61.7 4.3 2000 163,276 330,762 24,499 31.5 63.8 4.7 2005 160,597 363,912 28,641 29.0 65.8 5.2 2010 157,379 395,256 32,403 26.9 67.6 5.5 2015 154,953 423,636 36,837 25.2 68.8 6.0 2020 154,321 445,453 44,197 24.0 69.2 6.9 2025 154,426 460,925 54,588 23.1 68.8 8.1 2030 153,561 472,687 66,191 22.2 68.3 9.6 2035 152,021 480,027 79,079 21.4 67.5 11.1 2040 150,319 484,174 91,609 20.7 66.7 12.6 2045 149,285 485,418 102,931 20.2 65.8 14.0 2050 148,840 484,041 113,187 19.9 64.9 15.2 Low Mortality Scenario 1995 163,217 296,477 20,768 34.0 61.7 4.3 2000 163,276 330,762 24,499 31.5 63.8 4.7 2005 161,587 365,381 29,165 29.1 65.7 5.2 2010 159,787 399,398 34,057 26.9 67.3 5.7 2015 158,929 431,143 40,399 25.2 68.4 6.4 2020 159,382 456,890 50,542 23.9 68.5 7.6 2025 160,215 476,485 64,754 22.8 67.9 9.2 Table A12
Summary of Alternative Population Projections Southeast Asia High Fertility Scenario
Population (1000s) Percentage Year 0-14 15-64 65+ 0-14 15-64 65+ Standard Mortality Scenario 1995 163,217 296,477 20,768 34.0 61.7 4.3 2000 166,674 330,762 24,499 31.9 63.4 4.7 2005 171,938 364,308 28,810 30.4 64.5 5.1 2010 179,623 396,836 32,956 29.5 65.1 5.4 2015 185,881 430,117 38,004 28.4 65.8 5.8 2020 191,136 461,309 46,347 27.4 66.0 6.6 2025 197,073 489,074 58,241 26.5 65.7 7.8 2030 204,214 514,222 71,918 25.8 65.1 9.1 2035 212,341 536,310 87,409 25.4 64.1 10.5 2040 220,406 557,637 102,944 25.0 63.3 11.7 2045 228,525 579,099 117,510 24.7 62.6 12.7 2050 236,965 600,849 131,080 24.5 62.0 13.5 High Mortality Scenario 1995 163,217 296,477 20,768 34.0 61.7 4.3 2000 166,674 330,762 24,499 31.9 63.4 4.7 2005 171,527 363,912 28,641 30.4 64.5 5.1 2010 178,480 395,256 32,403 29.4 65.2 5.3 2015 183,862 426,942 36,837 28.4 65.9 5.7 2020 188,366 456,127 44,197 27.4 66.2 6.4 2025 193,490 481,577 54,588 26.5 66.0 7.5 2030 199,795 504,249 66,191 25.9 65.5 8.6 2035 206,945 523,850 79,079 25.6 64.7 9.8 2040 213,939 542,707 91,609 25.2 64.0 10.8 2045 220,964 561,677 102,931 25.0 63.4 11.6 2050 228,374 580,806 113,187 24.8 63.0 12.3 Low Mortality Scenario 1995 163,217 296,477 20,768 34.0 61.7 4.3 2000 166,674 330,762 24,499 31.9 63.4 4.7 2005 172,641 365,381 29,165 30.4 64.4 5.1 2010 181,317 399,426 34,057 29.5 65.0 5.5 2015 188,673 434,605 40,399 28.4 65.5 6.1 2020 194,676 468,059 50,542 27.3 65.6 7.1 2025 200,962 498,236 64,754 26.3 65.2 8.5 Table A13
Summary of Alternative Population Projections Southeast Asia Low Fertility Scenario
Population (1000s) Percentage Year 0-14 15-64 65+ 0-14 15-64 65+ Standard Mortality Scenario 1995 163,217 296,477 20,768 34.0 61.7 4.3 2000 159,730 330,762 24,499 31.0 64.2 4.8 2005 149,405 364,308 28,810 27.5 67.2 5.3 2010 136,502 396,836 32,956 24.1 70.1 5.8 2015 126,698 423,294 38,004 21.5 72.0 6.5 2020 121,751 439,081 46,347 20.1 72.3 7.6 2025 118,278 446,428 58,241 19.0 71.7 9.3 2030 113,048 448,858 71,918 17.8 70.8 11.3 2035 106,457 445,528 87,409 16.6 69.7 13.7 2040 99,666 437,263 102,944 15.6 68.3 16.1 2045 94,152 423,943 117,510 14.8 66.7 18.5 2050 89,816 405,987 131,080 14.3 64.8 20.9 High Mortality Scenario 1995 163,217 296,477 20,768 34.0 61.7 4.3 2000 159,730 330,762 24,499 31.0 64.2 4.8 2005 149,085 363,912 28,641 27.5 67.2 5.3 2010 135,706 395,256 32,403 24.1 70.2 5.8 2015 125,417 420,188 36,837 21.5 72.1 6.3 2020 120,104 434,216 44,197 20.1 72.5 7.4 2025 116,245 439,721 54,588 19.0 72.0 8.9 2030 110,734 440,385 66,191 17.9 71.3 10.7 2035 103,959 435,502 79,079 16.8 70.4 12.8 2040 97,061 425,994 91,609 15.8 69.3 14.9 2045 91,449 411,824 102,931 15.1 67.9 17.0 2050 87,022 393,404 113,187 14.7 66.3 19.1 Low Mortality Scenario 1995 163,217 296,477 20,768 34.0 61.7 4.3 2000 159,730 330,762 24,499 31.0 64.2 4.8 2005 149,936 365,381 29,165 27.5 67.1 5.4 2010 137,639 399,364 34,057 24.1 69.9 6.0 2015 128,499 427,491 40,399 21.5 71.7 6.8 2020 123,865 445,022 50,542 20.0 71.8 8.2 2025 120,341 453,934 64,754 18.8 71.0 10.1 Table A14
Summary of Aging, Alternative Mortality Scenarios Southeast Asia
Percentage 65 and older Old Age Dependency Ratio Year Medium High Low Medium High Low Medium Fertility Scenario 1995 4.3 4.3 4.3 0.070 0.070 0.070 2000 4.7 4.7 4.7 0.074 0.074 0.074 2005 5.2 5.2 5.2 0.079 0.079 0.080 2010 5.6 5.5 5.7 0.083 0.082 0.085 2015 6.1 6.0 6.4 0.089 0.087 0.094 2020 7.1 6.9 7.6 0.103 0.099 0.111 2025 8.5 8.1 9.2 0.124 0.118 0.136 2030 10.1 9.6 0.149 0.140 2035 11.9 11.1 0.178 0.165 2040 13.6 12.6 0.207 0.189 2045 15.2 14.0 0.235 0.212 2050 16.7 15.2 0.262 0.234 High Fertility Scenario 1995 4.3 4.3 4.3 0.070 0.070 0.070 2000 4.7 4.7 4.7 0.074 0.074 0.074 2005 5.1 5.1 5.1 0.079 0.079 0.080 2010 5.4 5.3 5.5 0.083 0.082 0.085 2015 5.8 5.7 6.1 0.088 0.086 0.093 2020 6.6 6.4 7.1 0.100 0.097 0.108 2025 7.8 7.5 8.5 0.119 0.113 0.130 2030 9.1 8.6 0.140 0.131 2035 10.5 9.8 0.163 0.151 2040 11.7 10.8 0.185 0.169 2045 12.7 11.6 0.203 0.183 2050 13.5 12.3 0.218 0.195 Low Fertility Scenario 1995 4.3 4.3 4.3 0.070 0.070 0.070 2000 4.8 4.8 4.8 0.074 0.074 0.074 2005 5.3 5.3 5.4 0.079 0.079 0.080 2010 5.8 5.8 6.0 0.083 0.082 0.085 2015 6.5 6.3 6.8 0.090 0.088 0.095 2020 7.6 7.4 8.2 0.106 0.102 0.114 2025 9.3 8.9 10.1 0.130 0.124 0.143 2030 11.3 10.7 0.160 0.150 2035 13.7 12.8 0.196 0.182 2040 16.1 14.9 0.235 0.215 2045 18.5 17.0 0.277 0.250 2050 20.9 19.1 0.323 0.288 Table A15
Total Dependency Ratio Southeast Asia
Fertility: High Med Low High Med Low High Med Low Mortality: High High High Med Med Med Low Low Low
1995 0.621 0.621 0.621 0.621 0.621 0.621 0.621 0.621 0.621 2000 0.578 0.568 0.557 0.578 0.568 0.557 0.578 0.568 0.557 2005 0.550 0.520 0.488 0.551 0.521 0.489 0.552 0.522 0.490 2010 0.534 0.480 0.425 0.536 0.482 0.427 0.539 0.485 0.430 2015 0.517 0.453 0.386 0.521 0.456 0.389 0.531 0.462 0.395 2020 0.510 0.446 0.378 0.515 0.450 0.383 0.537 0.459 0.392 2025 0.515 0.453 0.389 0.522 0.460 0.395 0.558 0.472 0.408 2030 0.527 0.465 0.402 0.537 0.475 0.412 2035 0.546 0.481 0.420 0.559 0.495 0.435 2040 0.563 0.500 0.443 0.580 0.518 0.463 2045 0.577 0.520 0.472 0.598 0.543 0.499 2050 0.588 0.541 0.509 0.613 0.570 0.544 Table A16
Summary of Alternative Population Projections South Asia Medium Fertility Scenario
Population (1000s) Percentage Year 0-14 15-64 65+ 0-14 15-64 65+ Standard Mortality Scenario 1995 505,326 801,284 58,563 37.0 58.7 4.3 2000 518,008 904,313 68,435 34.7 60.7 4.6 2005 519,506 1,014,732 79,282 32.2 62.9 4.9 2010 516,201 1,122,582 90,645 29.8 64.9 5.2 2015 509,415 1,224,264 105,618 27.7 66.6 5.7 2020 504,325 1,315,868 127,366 25.9 67.6 6.5 2025 501,584 1,392,773 155,438 24.5 67.9 7.6 2030 498,884 1,456,953 187,368 23.3 68.0 8.7 2035 493,628 1,512,478 222,393 22.1 67.9 10.0 2040 489,708 1,555,132 259,757 21.2 67.5 11.3 2045 489,273 1,585,218 297,782 20.6 66.8 12.6 2050 489,108 1,596,509 344,206 20.1 65.7 14.2 High Mortality Scenario 1995 505,326 801,284 58,563 37.0 58.7 4.3 2000 518,008 904,313 68,435 34.7 60.7 4.6 2005 518,173 1,013,331 78,746 32.2 62.9 4.9 2010 511,300 1,119,354 88,812 29.7 65.1 5.2 2015 501,478 1,216,963 101,887 27.5 66.9 5.6 2020 494,649 1,301,347 120,828 25.8 67.9 6.3 2025 489,952 1,368,888 144,931 24.5 68.3 7.2 2030 484,446 1,423,219 171,878 23.3 68.4 8.3 2035 475,621 1,468,865 200,881 22.2 68.5 9.4 2040 468,317 1,501,696 231,542 21.3 68.2 10.5 2045 465,056 1,522,227 261,932 20.7 67.7 11.6 2050 463,030 1,525,102 299,410 20.2 66.7 13.1 Low Mortality Scenario 1995 505,326 801,284 58,563 37.0 58.7 4.3 2000 518,008 904,313 68,435 34.7 60.7 4.6 2005 522,993 1,016,394 80,499 32.3 62.7 5.0 2010 522,721 1,129,388 94,169 29.9 64.7 5.4 2015 519,481 1,238,680 112,562 27.8 66.2 6.0 2020 517,597 1,338,939 138,659 25.9 67.1 6.9 2025 517,101 1,423,957 172,301 24.5 67.4 8.2 Table A17
Summary of Alternative Population Projections South Asia High Fertility Scenario
Population (1000s) Percentage Year 0-14 15-64 65+ 0-14 15-64 65+ Standard Mortality Scenario 1995 505,326 801,284 58,563 37.0 58.7 4.3 2000 526,254 904,313 68,435 35.1 60.3 4.6 2005 544,528 1,014,732 79,282 33.2 61.9 4.8 2010 567,444 1,122,582 90,645 31.9 63.0 5.1 2015 586,283 1,232,275 105,618 30.5 64.0 5.5 2020 604,427 1,340,354 127,366 29.2 64.7 6.1 2025 623,739 1,443,158 155,438 28.1 64.9 7.0 2030 645,325 1,540,630 187,368 27.2 64.9 7.9 2035 667,896 1,635,562 222,393 26.4 64.8 8.8 2040 692,821 1,725,785 259,757 25.9 64.4 9.7 2045 721,161 1,812,918 297,782 25.5 64.0 10.5 2050 750,369 1,890,694 344,206 25.1 63.3 11.5 High Mortality Scenario 1995 505,326 801,284 58,563 37.0 58.7 4.3 2000 526,254 904,313 68,435 35.1 60.3 4.6 2005 543,071 1,013,331 78,746 33.2 62.0 4.8 2010 561,928 1,119,354 88,812 31.7 63.2 5.0 2015 577,040 1,224,859 101,887 30.3 64.3 5.4 2020 592,726 1,325,386 120,828 29.1 65.0 5.9 2025 609,152 1,418,052 144,931 28.0 65.3 6.7 2030 626,480 1,504,569 171,878 27.2 65.3 7.5 2035 643,246 1,588,117 200,881 26.4 65.3 8.3 2040 662,223 1,666,288 231,542 25.9 65.1 9.0 2045 685,130 1,740,692 261,932 25.5 64.8 9.7 2050 710,021 1,805,682 299,410 25.2 64.1 10.6 Low Mortality Scenario 1995 505,326 801,284 58,563 37.0 58.7 4.3 2000 526,254 904,313 68,435 35.1 60.3 4.6 2005 548,287 1,016,394 80,499 33.3 61.8 4.9 2010 574,738 1,129,435 94,169 32.0 62.8 5.2 2015 597,965 1,246,906 112,562 30.5 63.7 5.8 2020 620,374 1,364,185 138,659 29.2 64.3 6.5 2025 643,068 1,475,995 172,301 28.1 64.4 7.5 Table A18
Summary of Alternative Population Projections South Asia Low Fertility Scenario
Population (1000s) Percentage Year 0-14 15-64 65+ 0-14 15-64 65+ Standard Mortality Scenario 1995 505,326 801,284 58,563 37.0 58.7 4.3 2000 509,282 904,313 68,435 34.4 61.0 4.6 2005 493,889 1,014,732 79,282 31.1 63.9 5.0 2010 463,980 1,122,582 90,645 27.7 66.9 5.4 2015 431,302 1,215,813 105,618 24.6 69.4 6.0 2020 403,926 1,290,805 127,366 22.2 70.8 7.0 2025 382,797 1,341,441 155,438 20.4 71.4 8.3 2030 363,310 1,371,629 187,368 18.9 71.4 9.7 2035 339,959 1,388,545 222,393 17.4 71.2 11.4 2040 318,346 1,386,890 259,757 16.2 70.6 13.2 2045 301,439 1,366,681 297,782 15.3 69.5 15.1 2050 286,272 1,321,962 344,206 14.7 67.7 17.6 High Mortality Scenario 1995 505,326 801,284 58,563 37.0 58.7 4.3 2000 509,282 904,313 68,435 34.4 61.0 4.6 2005 492,697 1,013,331 78,746 31.1 63.9 5.0 2010 459,743 1,119,354 88,812 27.6 67.1 5.3 2015 424,766 1,208,618 101,887 24.5 69.7 5.9 2020 396,328 1,276,769 120,828 22.1 71.2 6.7 2025 374,049 1,318,843 144,931 20.4 71.8 7.9 2030 352,914 1,340,340 171,878 18.9 71.9 9.2 2035 327,747 1,348,884 200,881 17.5 71.8 10.7 2040 304,693 1,339,521 231,542 16.2 71.4 12.3 2045 286,810 1,312,635 261,932 15.4 70.5 14.1 2050 271,325 1,263,308 299,410 14.8 68.9 16.3 Low Mortality Scenario 1995 505,326 801,284 58,563 37.0 58.7 4.3 2000 509,282 904,313 68,435 34.4 61.0 4.6 2005 497,069 1,016,394 80,499 31.2 63.8 5.1 2010 469,656 1,129,339 94,169 27.7 66.7 5.6 2015 439,691 1,229,949 112,562 24.7 69.0 6.3 2020 414,492 1,312,951 138,659 22.2 70.4 7.4 2025 394,549 1,370,669 172,301 20.4 70.7 8.9 Table A19
Summary of Aging, Alternative Mortality Scenarios South Asia
Percentage 65 and older Old Age Dependency Ratio Year Medium High Low Medium High Low Medium Fertility Scenario 1995 4.3 4.3 4.3 0.073 0.073 0.073 2000 4.6 4.6 4.6 0.076 0.076 0.076 2005 4.9 4.9 5.0 0.078 0.078 0.079 2010 5.2 5.2 5.4 0.081 0.079 0.083 2015 5.7 5.6 6.0 0.086 0.084 0.091 2020 6.5 6.3 6.9 0.097 0.093 0.104 2025 7.6 7.2 8.2 0.112 0.106 0.121 2030 8.7 8.3 0.129 0.121 2035 10.0 9.4 0.147 0.137 2040 11.3 10.5 0.167 0.154 2045 12.6 11.6 0.188 0.172 2050 14.2 13.1 0.216 0.196 High Fertility Scenario 1995 4.3 4.3 4.3 0.073 0.073 0.073 2000 4.6 4.6 4.6 0.076 0.076 0.076 2005 4.8 4.8 4.9 0.078 0.078 0.079 2010 5.1 5.0 5.2 0.081 0.079 0.083 2015 5.5 5.4 5.8 0.086 0.083 0.090 2020 6.1 5.9 6.5 0.095 0.091 0.102 2025 7.0 6.7 7.5 0.108 0.102 0.117 2030 7.9 7.5 0.122 0.114 2035 8.8 8.3 0.136 0.126 2040 9.7 9.0 0.151 0.139 2045 10.5 9.7 0.164 0.150 2050 11.5 10.6 0.182 0.166 Low Fertility Scenario 1995 4.3 4.3 4.3 0.073 0.073 0.073 2000 4.6 4.6 4.6 0.076 0.076 0.076 2005 5.0 5.0 5.1 0.078 0.078 0.079 2010 5.4 5.3 5.6 0.081 0.079 0.083 2015 6.0 5.9 6.3 0.087 0.084 0.092 2020 7.0 6.7 7.4 0.099 0.095 0.106 2025 8.3 7.9 8.9 0.116 0.110 0.126 2030 9.7 9.2 0.137 0.128 2035 11.4 10.7 0.160 0.149 2040 13.2 12.3 0.187 0.173 2045 15.1 14.1 0.218 0.200 2050 17.6 16.3 0.260 0.237 Table A20
Total Dependency Ratio South Asia
Fertility: High Med Low High Med Low High Med Low Mortality: High High High Med Med Med Low Low Low
1995 0.704 0.704 0.704 0.704 0.704 0.704 0.704 0.704 0.704 2000 0.658 0.648 0.639 0.658 0.648 0.639 0.658 0.648 0.639 2005 0.614 0.589 0.564 0.615 0.590 0.565 0.619 0.594 0.568 2010 0.581 0.536 0.490 0.586 0.541 0.494 0.592 0.546 0.499 2015 0.554 0.496 0.436 0.561 0.502 0.442 0.574 0.510 0.449 2020 0.538 0.473 0.405 0.546 0.480 0.412 0.567 0.490 0.421 2025 0.532 0.464 0.394 0.540 0.472 0.401 0.573 0.484 0.414 2030 0.531 0.461 0.392 0.540 0.471 0.401 2035 0.532 0.461 0.392 0.544 0.473 0.405 2040 0.536 0.466 0.400 0.552 0.482 0.417 2045 0.544 0.478 0.418 0.562 0.496 0.438 2050 0.559 0.500 0.452 0.579 0.522 0.477 Table A21
Summary of Alternative Population Projections South Korea Medium Fertility Scenario
Population (1000s) Percentage Year 0-14 15-64 65+ 0-14 15-64 65+ Standard Mortality Scenario 1995 10,540 31,882 2,527 23.4 70.9 5.6 2000 10,068 33,623 3,152 21.5 71.8 6.7 2005 10,118 34,496 3,934 20.8 71.1 8.1 2010 9,882 35,455 4,639 19.8 70.9 9.3 2015 9,558 36,106 5,387 18.7 70.7 10.6 2020 9,213 36,314 6,366 17.8 70.0 12.3 2025 8,956 35,557 8,020 17.0 67.7 15.3 2030 8,827 34,506 9,564 16.7 65.2 18.1 2035 8,733 33,379 10,865 16.5 63.0 20.5 2040 8,592 32,083 12,020 16.3 60.9 22.8 2045 8,404 31,205 12,482 16.1 59.9 24.0 2050 8,209 30,401 12,656 16.0 59.3 24.7 High Mortality Scenario 1995 10,540 31,882 2,527 23.4 70.9 5.6 2000 10,068 33,623 3,152 21.5 71.8 6.7 2005 10,114 34,462 3,908 20.9 71.1 8.1 2010 9,872 35,362 4,562 19.8 71.0 9.2 2015 9,539 35,936 5,234 18.8 70.9 10.3 2020 9,178 36,017 6,097 17.9 70.2 11.9 2025 8,898 35,120 7,563 17.3 68.1 14.7 2030 8,746 33,955 8,870 17.0 65.8 17.2 2035 8,634 32,754 9,887 16.8 63.9 19.3 2040 8,479 31,442 10,757 16.7 62.0 21.2 2045 8,282 30,557 10,980 16.6 61.3 22.0 2050 8,084 29,772 10,966 16.6 61.0 22.5 Low Mortality Scenario 1995 10,540 31,882 2,527 23.4 70.9 5.6 2000 10,068 33,623 3,152 21.5 71.8 6.7 2005 10,124 34,553 3,981 20.8 71.0 8.2 2010 9,907 35,642 4,782 19.7 70.8 9.5 2015 9,609 36,468 5,687 18.6 70.4 11.0 2020 9,279 36,790 6,884 17.5 69.5 13.0 2025 9,027 36,060 8,817 16.7 66.9 16.4 Table A22
Summary of Alternative Population Projections South Korea High Fertility Scenario
Population (1000s) Percentage Year 0-14 15-64 65+ 0-14 15-64 65+ Standard Mortality Scenario 1995 10,540 31,882 2,527 23.4 70.9 5.6 2000 10,212 33,623 3,152 21.7 71.6 6.7 2005 10,547 34,496 3,934 21.5 70.4 8.0 2010 10,632 35,455 4,639 21.0 69.9 9.1 2015 10,508 36,249 5,387 20.2 69.5 10.3 2020 10,272 36,742 6,366 19.2 68.8 11.9 2025 10,297 36,303 8,020 18.9 66.5 14.7 2030 10,618 35,595 9,564 19.0 63.8 17.1 2035 11,039 34,859 10,865 19.4 61.4 19.1 2040 11,245 34,161 12,020 19.6 59.5 20.9 2045 11,263 34,070 12,482 19.5 58.9 21.6 2050 11,283 34,166 12,656 19.4 58.8 21.8 High Mortality Scenario 1995 10,540 31,882 2,527 23.4 70.9 5.6 2000 10,212 33,623 3,152 21.7 71.6 6.7 2005 10,543 34,462 3,908 21.6 70.5 8.0 2010 10,621 35,362 4,562 21.0 70.0 9.0 2015 10,486 36,078 5,234 20.2 69.7 10.1 2020 10,233 36,441 6,097 19.4 69.1 11.6 2025 10,231 35,860 7,563 19.1 66.8 14.1 2030 10,522 35,033 8,870 19.3 64.4 16.3 2035 10,914 34,217 9,887 19.8 62.2 18.0 2040 11,097 33,494 10,757 20.0 60.5 19.4 2045 11,101 33,384 10,980 20.0 60.2 19.8 2050 11,114 33,482 10,966 20.0 60.3 19.7 Low Mortality Scenario 1995 10,540 31,882 2,527 23.4 70.9 5.6 2000 10,212 33,623 3,152 21.7 71.6 6.7 2005 10,554 34,553 3,981 21.5 70.4 8.1 2010 10,659 35,642 4,782 20.9 69.8 9.4 2015 10,563 36,613 5,687 20.0 69.3 10.8 2020 10,344 37,220 6,884 19.0 68.4 12.6 2025 10,377 36,810 8,817 18.5 65.7 15.7 Table A23
Summary of Alternative Population Projections South Korea Low Fertility Scenario
Population (1000s) Percentage Year 0-14 15-64 65+ 0-14 15-64 65+ Standard Mortality Scenario 1995 10,540 31,882 2,527 23.4 70.9 5.6 2000 9,925 33,623 3,152 21.3 72.0 6.7 2005 9,609 34,496 3,934 20.0 71.8 8.2 2010 8,907 35,455 4,639 18.2 72.4 9.5 2015 8,182 35,963 5,387 16.5 72.6 10.9 2020 7,577 35,808 6,366 15.2 72.0 12.8 2025 7,099 34,585 8,020 14.3 69.6 16.1 2030 6,728 32,994 9,564 13.7 66.9 19.4 2035 6,380 31,245 10,865 13.2 64.4 22.4 2040 5,986 29,267 12,020 12.7 61.9 25.4 2045 5,597 27,611 12,482 12.2 60.4 27.3 2050 5,251 25,940 12,656 12.0 59.1 28.9 High Mortality Scenario 1995 10,540 31,882 2,527 23.4 70.9 5.6 2000 9,925 33,623 3,152 21.3 72.0 6.7 2005 9,605 34,462 3,908 20.0 71.8 8.1 2010 8,899 35,362 4,562 18.2 72.4 9.3 2015 8,168 35,794 5,234 16.6 72.8 10.6 2020 7,549 35,514 6,097 15.4 72.2 12.4 2025 7,052 34,157 7,563 14.5 70.0 15.5 2030 6,664 32,459 8,870 13.9 67.6 18.5 2035 6,305 30,645 9,887 13.5 65.4 21.1 2040 5,906 28,662 10,757 13.0 63.2 23.7 2045 5,515 27,014 10,980 12.7 62.1 25.2 2050 5,169 25,377 10,966 12.5 61.1 26.4 Low Mortality Scenario 1995 10,540 31,882 2,527 23.4 70.9 5.6 2000 9,925 33,623 3,152 21.3 72.0 6.7 2005 9,613 34,553 3,981 20.0 71.8 8.3 2010 8,928 35,641 4,782 18.1 72.2 9.7 2015 8,228 36,324 5,687 16.4 72.3 11.3 2020 7,634 36,280 6,884 15.0 71.4 13.6 2025 7,156 35,083 8,817 14.0 68.7 17.3 Table A24
Summary of Aging, Alternative Mortality Scenarios South Korea
Percentage 65 and older Old Age Dependency Ratio Year Medium High Low Medium High Low Medium Fertility Scenario 1995 5.6 5.6 5.6 0.079 0.079 0.079 2000 6.7 6.7 6.7 0.094 0.094 0.094 2005 8.1 8.1 8.2 0.114 0.113 0.115 2010 9.3 9.2 9.5 0.131 0.129 0.134 2015 10.6 10.3 11.0 0.149 0.146 0.156 2020 12.3 11.9 13.0 0.175 0.169 0.187 2025 15.3 14.7 16.4 0.226 0.215 0.245 2030 18.1 17.2 0.277 0.261 2035 20.5 19.3 0.326 0.302 2040 22.8 21.2 0.375 0.342 2045 24.0 22.0 0.400 0.359 2050 24.7 22.5 0.416 0.368 High Fertility Scenario 1995 5.6 5.6 5.6 0.079 0.079 0.079 2000 6.7 6.7 6.7 0.094 0.094 0.094 2005 8.0 8.0 8.1 0.114 0.113 0.115 2010 9.1 9.0 9.4 0.131 0.129 0.134 2015 10.3 10.1 10.8 0.149 0.145 0.155 2020 11.9 11.6 12.6 0.173 0.167 0.185 2025 14.7 14.1 15.7 0.221 0.211 0.240 2030 17.1 16.3 0.269 0.253 2035 19.1 18.0 0.312 0.289 2040 20.9 19.4 0.352 0.321 2045 21.6 19.8 0.366 0.329 2050 21.8 19.7 0.370 0.328 Low Fertility Scenario 1995 5.6 5.6 5.6 0.079 0.079 0.079 2000 6.7 6.7 6.7 0.094 0.094 0.094 2005 8.2 8.1 8.3 0.114 0.113 0.115 2010 9.5 9.3 9.7 0.131 0.129 0.134 2015 10.9 10.6 11.3 0.150 0.146 0.157 2020 12.8 12.4 13.6 0.178 0.172 0.190 2025 16.1 15.5 17.3 0.232 0.221 0.251 2030 19.4 18.5 0.290 0.273 2035 22.4 21.1 0.348 0.323 2040 25.4 23.7 0.411 0.375 2045 27.3 25.2 0.452 0.406 2050 28.9 26.4 0.488 0.432 Table A25
Total Dependency Ratio South Korea
Fertility: High Med Low High Med Low High Med Low Mortality: High High High Med Med Med Low Low Low
1995 0.410 0.410 0.410 0.410 0.410 0.410 0.410 0.410 0.410 2000 0.397 0.393 0.389 0.397 0.393 0.389 0.397 0.393 0.389 2005 0.419 0.407 0.392 0.420 0.407 0.393 0.421 0.408 0.393 2010 0.429 0.408 0.381 0.431 0.410 0.382 0.433 0.412 0.385 2015 0.436 0.411 0.374 0.438 0.414 0.377 0.446 0.419 0.383 2020 0.448 0.424 0.384 0.453 0.429 0.389 0.468 0.439 0.400 2025 0.496 0.469 0.428 0.505 0.477 0.437 0.532 0.495 0.455 2030 0.554 0.519 0.479 0.567 0.533 0.494 2035 0.608 0.565 0.528 0.628 0.587 0.552 2040 0.652 0.612 0.581 0.681 0.642 0.615 2045 0.661 0.630 0.611 0.697 0.669 0.655 2050 0.659 0.640 0.636 0.701 0.686 0.690 Table A26
Summary of Alternative Population Projections The Philippines Medium Fertility Scenario
Population (1000s) Percentage Year 0-14 15-64 65+ 0-14 15-64 65+ Standard Mortality Scenario 1995 26,213 39,779 2,362 38.3 58.2 3.5 2000 27,870 45,339 2,758 36.7 59.7 3.6 2005 28,752 51,403 3,295 34.5 61.6 3.9 2010 28,990 57,632 3,923 32.0 63.7 4.3 2015 27,971 63,887 4,874 28.9 66.0 5.0 2020 26,694 69,597 6,114 26.1 68.0 6.0 2025 26,222 74,242 7,788 24.2 68.6 7.2 2030 26,720 77,771 9,534 23.4 68.2 8.4 2035 27,436 80,315 11,565 23.0 67.3 9.7 2040 27,396 82,929 13,534 22.1 67.0 10.9 2045 26,815 84,790 16,032 21.0 66.4 12.6 2050 26,253 86,082 18,559 20.1 65.8 14.2 High Mortality Scenario 1995 26,213 39,779 2,362 38.3 58.2 3.5 2000 27,870 45,339 2,758 36.7 59.7 3.6 2005 28,705 51,328 3,276 34.5 61.6 3.9 2010 28,842 57,339 3,858 32.0 63.7 4.3 2015 27,711 63,342 4,730 28.9 66.1 4.9 2020 26,361 68,775 5,839 26.1 68.1 5.8 2025 25,823 73,114 7,298 24.3 68.8 6.9 2030 26,257 76,311 8,745 23.6 68.6 7.9 2035 26,895 78,554 10,396 23.2 67.8 9.0 2040 26,787 80,869 11,932 22.4 67.6 10.0 2045 26,167 82,507 13,882 21.4 67.3 11.3 2050 25,585 83,605 15,802 20.5 66.9 12.6 Low Mortality Scenario 1995 26,213 39,779 2,362 38.3 58.2 3.5 2000 27,870 45,339 2,758 36.7 59.7 3.6 2005 28,847 51,597 3,335 34.4 61.6 4.0 2010 29,199 58,038 4,053 32.0 63.6 4.4 2015 28,290 64,541 5,190 28.9 65.8 5.3 2020 27,093 70,515 6,691 26.0 67.6 6.4 2025 26,620 75,423 8,736 24.0 68.1 7.9 Table A27
Summary of Alternative Population Projections The Philippines High Fertility Scenario
Population (1000s) Percentage Year 0-14 15-64 65+ 0-14 15-64 65+ Standard Mortality Scenario 1995 26,213 39,779 2,362 38.3 58.2 3.5 2000 28,417 45,339 2,758 37.1 59.3 3.6 2005 30,148 51,403 3,295 35.5 60.6 3.9 2010 31,598 57,632 3,923 33.9 61.9 4.2 2015 31,687 64,428 4,874 31.4 63.8 4.8 2020 31,659 70,982 6,114 29.1 65.3 5.6 2025 32,448 76,832 7,788 27.7 65.6 6.7 2030 34,213 82,000 9,534 27.2 65.2 7.6 2035 36,204 86,627 11,565 26.9 64.5 8.6 2040 37,529 91,694 13,534 26.3 64.2 9.5 2045 38,349 96,446 16,032 25.4 63.9 10.6 2050 39,235 101,071 18,559 24.7 63.6 11.7 High Mortality Scenario 1995 26,213 39,779 2,362 38.3 58.2 3.5 2000 28,417 45,339 2,758 37.1 59.3 3.6 2005 30,096 51,328 3,276 35.5 60.6 3.9 2010 31,429 57,339 3,858 33.9 61.9 4.2 2015 31,381 63,875 4,730 31.4 63.9 4.7 2020 31,247 70,135 5,839 29.1 65.4 5.4 2025 31,926 75,644 7,298 27.8 65.9 6.4 2030 33,579 80,417 8,745 27.4 65.5 7.1 2035 35,429 84,652 10,396 27.2 64.9 8.0 2040 36,605 89,299 11,932 26.6 64.8 8.7 2045 37,294 93,670 13,882 25.7 64.7 9.6 2050 38,066 97,897 15,802 25.1 64.5 10.4 Low Mortality Scenario 1995 26,213 39,779 2,362 38.3 58.2 3.5 2000 28,417 45,339 2,758 37.1 59.3 3.6 2005 30,253 51,597 3,335 35.5 60.6 3.9 2010 31,841 58,046 4,053 33.9 61.8 4.3 2015 32,075 65,116 5,190 31.3 63.6 5.1 2020 32,169 71,994 6,691 29.0 64.9 6.0 2025 33,002 78,198 8,736 27.5 65.2 7.3 Table A28
Summary of Alternative Population Projections The Philippines Low Fertility Scenario
Population (1000s) Percentage Year 0-14 15-64 65+ 0-14 15-64 65+ Standard Mortality Scenario 1995 26,213 39,779 2,362 38.3 58.2 3.5 2000 27,596 45,339 2,758 36.5 59.9 3.6 2005 27,827 51,403 3,295 33.7 62.3 4.0 2010 26,950 57,632 3,923 30.5 65.1 4.4 2015 24,504 63,616 4,874 26.4 68.4 5.2 2020 21,803 68,680 6,114 22.6 71.1 6.3 2025 20,150 72,216 7,788 20.1 72.1 7.8 2030 19,774 74,055 9,534 19.1 71.6 9.2 2035 19,703 74,540 11,565 18.6 70.5 10.9 2040 18,812 74,874 13,534 17.5 69.8 12.6 2045 17,387 74,187 16,032 16.2 68.9 14.9 2050 16,005 72,652 18,559 14.9 67.8 17.3 High Mortality Scenario 1995 26,213 39,779 2,362 38.3 58.2 3.5 2000 27,596 45,339 2,758 36.5 59.9 3.6 2005 27,784 51,328 3,276 33.7 62.3 4.0 2010 26,819 57,339 3,858 30.5 65.1 4.4 2015 24,291 63,075 4,730 26.4 68.5 5.1 2020 21,555 67,876 5,839 22.6 71.2 6.1 2025 19,876 71,138 7,298 20.2 72.4 7.4 2030 19,469 72,709 8,745 19.3 72.0 8.7 2035 19,359 72,984 10,396 18.8 71.0 10.1 2040 18,456 73,136 11,932 17.8 70.6 11.5 2045 17,058 72,371 13,882 16.5 70.1 13.4 2050 15,724 70,820 15,802 15.4 69.2 15.4 Low Mortality Scenario 1995 26,213 39,779 2,362 38.3 58.2 3.5 2000 27,596 45,339 2,758 36.5 59.9 3.6 2005 27,914 51,597 3,335 33.7 62.3 4.0 2010 27,119 58,033 4,053 30.4 65.1 4.5 2015 24,740 64,236 5,190 26.3 68.2 5.5 2020 22,085 69,476 6,691 22.5 70.7 6.8 2025 20,391 73,138 8,736 19.9 71.5 8.5 Table A29
Summary of Aging, Alternative Mortality Scenarios The Philippines
Percentage 65 and older Old Age Dependency Ratio Year Medium High Low Medium High Low Medium Fertility Scenario 1995 3.5 3.5 3.5 0.059 0.059 0.059 2000 3.6 3.6 3.6 0.061 0.061 0.061 2005 3.9 3.9 4.0 0.064 0.064 0.065 2010 4.3 4.3 4.4 0.068 0.067 0.070 2015 5.0 4.9 5.3 0.076 0.075 0.080 2020 6.0 5.8 6.4 0.088 0.085 0.095 2025 7.2 6.9 7.9 0.105 0.100 0.116 2030 8.4 7.9 0.123 0.115 2035 9.7 9.0 0.144 0.132 2040 10.9 10.0 0.163 0.148 2045 12.6 11.3 0.189 0.168 2050 14.2 12.6 0.216 0.189 High Fertility Scenario 1995 3.5 3.5 3.5 0.059 0.059 0.059 2000 3.6 3.6 3.6 0.061 0.061 0.061 2005 3.9 3.9 3.9 0.064 0.064 0.065 2010 4.2 4.2 4.3 0.068 0.067 0.070 2015 4.8 4.7 5.1 0.076 0.074 0.080 2020 5.6 5.4 6.0 0.086 0.083 0.093 2025 6.7 6.4 7.3 0.101 0.096 0.112 2030 7.6 7.1 0.116 0.109 2035 8.6 8.0 0.133 0.123 2040 9.5 8.7 0.148 0.134 2045 10.6 9.6 0.166 0.148 2050 11.7 10.4 0.184 0.161 Low Fertility Scenario 1995 3.5 3.5 3.5 0.059 0.059 0.059 2000 3.6 3.6 3.6 0.061 0.061 0.061 2005 4.0 4.0 4.0 0.064 0.064 0.065 2010 4.4 4.4 4.5 0.068 0.067 0.070 2015 5.2 5.1 5.5 0.077 0.075 0.081 2020 6.3 6.1 6.8 0.089 0.086 0.096 2025 7.8 7.4 8.5 0.108 0.103 0.119 2030 9.2 8.7 0.129 0.120 2035 10.9 10.1 0.155 0.142 2040 12.6 11.5 0.181 0.163 2045 14.9 13.4 0.216 0.192 2050 17.3 15.4 0.255 0.223 Table A30
Total Dependency Ratio The Philippines
Fertility: High Med Low High Med Low High Med Low Mortality: High High High Med Med Med Low Low Low
1995 0.718 0.718 0.718 0.718 0.718 0.718 0.718 0.718 0.718 2000 0.688 0.676 0.669 0.688 0.676 0.669 0.688 0.676 0.669 2005 0.650 0.623 0.605 0.651 0.623 0.605 0.651 0.624 0.606 2010 0.615 0.570 0.535 0.616 0.571 0.536 0.618 0.573 0.537 2015 0.565 0.512 0.460 0.567 0.514 0.462 0.577 0.519 0.466 2020 0.529 0.468 0.404 0.532 0.471 0.406 0.551 0.479 0.414 2025 0.519 0.453 0.382 0.524 0.458 0.387 0.553 0.469 0.398 2030 0.526 0.459 0.388 0.534 0.466 0.396 2035 0.541 0.475 0.408 0.551 0.486 0.419 2040 0.544 0.479 0.416 0.557 0.494 0.432 2045 0.546 0.485 0.428 0.564 0.505 0.450 2050 0.550 0.495 0.445 0.572 0.521 0.476 Table A31
Summary of Alternative Population Projections Thailand Medium Fertility Scenario
Population (1000s) Percentage Year 0-14 15-64 65+ 0-14 15-64 65+ Standard Mortality Scenario 1995 16,397 39,279 2,934 28.0 67.0 5.0 2000 15,489 42,334 3,576 25.2 68.9 5.8 2005 14,498 45,203 4,288 22.7 70.6 6.7 2010 14,343 47,215 4,953 21.6 71.0 7.4 2015 14,259 48,792 5,821 20.7 70.8 8.5 2020 14,061 49,744 7,170 19.8 70.1 10.1 2025 13,733 50,060 8,924 18.9 68.8 12.3 2030 13,372 49,743 10,914 18.1 67.2 14.7 2035 13,091 48,615 13,155 17.5 64.9 17.6 2040 12,899 47,251 15,030 17.2 62.9 20.0 2045 12,688 45,799 16,427 16.9 61.1 21.9 2050 12,431 44,680 17,077 16.8 60.2 23.0 High Mortality Scenario 1995 16,397 39,279 2,934 28.0 67.0 5.0 2000 15,489 42,334 3,576 25.2 68.9 5.8 2005 14,474 45,161 4,273 22.6 70.7 6.7 2010 14,275 46,956 4,883 21.6 71.0 7.4 2015 14,142 48,256 5,663 20.8 70.9 8.3 2020 13,907 48,934 6,842 20.0 70.2 9.8 2025 13,544 49,018 8,346 19.1 69.1 11.8 2030 13,173 48,571 10,058 18.3 67.6 14.0 2035 12,887 47,425 12,030 17.8 65.6 16.6 2040 12,688 46,105 13,681 17.5 63.6 18.9 2045 12,462 44,725 14,956 17.3 62.0 20.7 2050 12,193 43,671 15,542 17.1 61.2 21.8 Low Mortality Scenario 1995 16,397 39,279 2,934 28.0 67.0 5.0 2000 15,489 42,334 3,576 25.2 68.9 5.8 2005 14,544 45,399 4,337 22.6 70.6 6.7 2010 14,434 47,679 5,108 21.5 70.9 7.6 2015 14,389 49,456 6,137 20.6 70.7 8.8 2020 14,207 50,492 7,636 19.6 69.8 10.6 2025 13,868 50,824 9,541 18.7 68.5 12.9 Table A32
Summary of Alternative Population Projections Thailand High Fertility Scenario
Population (1000s) Percentage Year 0-14 15-64 65+ 0-14 15-64 65+ Standard Mortality Scenario 1995 16,397 39,279 2,934 28.0 67.0 5.0 2000 15,935 42,334 3,576 25.8 68.5 5.8 2005 15,660 45,203 4,288 24.0 69.4 6.6 2010 16,363 47,215 4,953 23.9 68.9 7.2 2015 16,683 49,231 5,821 23.3 68.6 8.1 2020 16,715 50,892 7,170 22.4 68.1 9.6 2025 16,973 52,056 8,924 21.8 66.8 11.4 2030 17,546 52,569 10,914 21.7 64.9 13.5 2035 18,230 52,362 13,155 21.8 62.5 15.7 2040 18,608 52,406 15,030 21.6 60.9 17.5 2045 18,914 52,682 16,427 21.5 59.8 18.7 2050 19,422 53,405 17,077 21.6 59.4 19.0 High Mortality Scenario 1995 16,397 39,279 2,934 28.0 67.0 5.0 2000 15,935 42,334 3,576 25.8 68.5 5.8 2005 15,633 45,161 4,273 24.0 69.4 6.6 2010 16,286 46,956 4,883 23.9 68.9 7.2 2015 16,548 48,692 5,663 23.3 68.7 8.0 2020 16,534 50,074 6,842 22.5 68.2 9.3 2025 16,740 50,997 8,346 22.0 67.0 11.0 2030 17,286 51,365 10,058 22.0 65.3 12.8 2035 17,947 51,121 12,030 22.1 63.0 14.8 2040 18,305 51,179 13,681 22.0 61.5 16.5 2045 18,579 51,492 14,956 21.9 60.6 17.6 2050 19,050 52,243 15,542 21.9 60.2 17.9 Low Mortality Scenario 1995 16,397 39,279 2,934 28.0 67.0 5.0 2000 15,935 42,334 3,576 25.8 68.5 5.8 2005 15,710 45,399 4,337 24.0 69.4 6.6 2010 16,467 47,679 5,108 23.8 68.8 7.4 2015 16,834 49,896 6,137 23.1 68.5 8.4 2020 16,889 51,645 7,636 22.2 67.8 10.0 2025 17,140 52,837 9,541 21.6 66.4 12.0 Table A33
Summary of Alternative Population Projections Thailand Low Fertility Scenario
Population (1000s) Percentage Year 0-14 15-64 65+ 0-14 15-64 65+ Standard Mortality Scenario 1995 16,397 39,279 2,934 28.0 67.0 5.0 2000 15,211 42,334 3,576 24.9 69.3 5.9 2005 13,640 45,203 4,288 21.6 71.6 6.8 2010 12,749 47,215 4,953 19.6 72.7 7.6 2015 12,041 48,518 5,821 18.1 73.1 8.8 2020 11,354 48,898 7,170 16.8 72.5 10.6 2025 10,589 48,486 8,924 15.6 71.3 13.1 2030 9,820 47,281 10,914 14.4 69.5 16.0 2035 9,170 45,109 13,155 13.6 66.9 19.5 2040 8,609 42,602 15,030 13.0 64.3 22.7 2045 8,076 39,884 16,427 12.5 61.9 25.5 2050 7,549 37,388 17,077 12.2 60.3 27.5 High Mortality Scenario 1995 16,397 39,279 2,934 28.0 67.0 5.0 2000 15,211 42,334 3,576 24.9 69.3 5.9 2005 13,618 45,161 4,273 21.6 71.6 6.8 2010 12,688 46,956 4,883 19.7 72.8 7.6 2015 11,942 47,983 5,663 18.2 73.2 8.6 2020 11,230 48,093 6,842 17.0 72.7 10.3 2025 10,443 47,456 8,346 15.8 71.6 12.6 2030 9,674 46,138 10,058 14.7 70.0 15.3 2035 9,027 43,968 12,030 13.9 67.6 18.5 2040 8,467 41,530 13,681 13.3 65.2 21.5 2045 7,933 38,910 14,956 12.8 63.0 24.2 2050 7,404 36,509 15,542 12.5 61.4 26.1 Low Mortality Scenario 1995 16,397 39,279 2,934 28.0 67.0 5.0 2000 15,211 42,334 3,576 24.9 69.3 5.9 2005 13,684 45,399 4,337 21.6 71.6 6.8 2010 12,829 47,679 5,108 19.6 72.7 7.8 2015 12,151 49,181 6,137 18.0 72.9 9.1 2020 11,472 49,640 7,636 16.7 72.2 11.1 2025 10,694 49,235 9,541 15.4 70.9 13.7 Table A34
Summary of Aging, Alternative Mortality Scenarios Thailand
Percentage 65 and older Old Age Dependency Ratio Year Medium High Low Medium High Low Medium Fertility Scenario 1995 5.0 5.0 5.0 0.075 0.075 0.075 2000 5.8 5.8 5.8 0.084 0.084 0.084 2005 6.7 6.7 6.7 0.095 0.095 0.096 2010 7.4 7.4 7.6 0.105 0.104 0.107 2015 8.5 8.3 8.8 0.119 0.117 0.124 2020 10.1 9.8 10.6 0.144 0.140 0.151 2025 12.3 11.8 12.9 0.178 0.170 0.188 2030 14.7 14.0 0.219 0.207 2035 17.6 16.6 0.271 0.254 2040 20.0 18.9 0.318 0.297 2045 21.9 20.7 0.359 0.334 2050 23.0 21.8 0.382 0.356 High Fertility Scenario 1995 5.0 5.0 5.0 0.075 0.075 0.075 2000 5.8 5.8 5.8 0.084 0.084 0.084 2005 6.6 6.6 6.6 0.095 0.095 0.096 2010 7.2 7.2 7.4 0.105 0.104 0.107 2015 8.1 8.0 8.4 0.118 0.116 0.123 2020 9.6 9.3 10.0 0.141 0.137 0.148 2025 11.4 11.0 12.0 0.171 0.164 0.181 2030 13.5 12.8 0.208 0.196 2035 15.7 14.8 0.251 0.235 2040 17.5 16.5 0.287 0.267 2045 18.7 17.6 0.312 0.290 2050 19.0 17.9 0.320 0.297 Low Fertility Scenario 1995 5.0 5.0 5.0 0.075 0.075 0.075 2000 5.9 5.9 5.9 0.084 0.084 0.084 2005 6.8 6.8 6.8 0.095 0.095 0.096 2010 7.6 7.6 7.8 0.105 0.104 0.107 2015 8.8 8.6 9.1 0.120 0.118 0.125 2020 10.6 10.3 11.1 0.147 0.142 0.154 2025 13.1 12.6 13.7 0.184 0.176 0.194 2030 16.0 15.3 0.231 0.218 2035 19.5 18.5 0.292 0.274 2040 22.7 21.5 0.353 0.329 2045 25.5 24.2 0.412 0.384 2050 27.5 26.1 0.457 0.426 Table A35
Total Dependency Ratio Thailand
Fertility: High Med Low High Med Low High Med Low Mortality: High High High Med Med Med Low Low Low
1995 0.492 0.492 0.492 0.492 0.492 0.492 0.492 0.492 0.492 2000 0.461 0.450 0.444 0.461 0.450 0.444 0.461 0.450 0.444 2005 0.441 0.415 0.396 0.441 0.416 0.397 0.442 0.416 0.397 2010 0.451 0.408 0.374 0.451 0.409 0.375 0.453 0.410 0.376 2015 0.456 0.410 0.367 0.457 0.412 0.368 0.464 0.415 0.372 2020 0.467 0.424 0.376 0.469 0.427 0.379 0.486 0.433 0.385 2025 0.492 0.447 0.396 0.497 0.453 0.402 0.525 0.461 0.411 2030 0.532 0.478 0.428 0.541 0.488 0.439 2035 0.586 0.525 0.479 0.599 0.540 0.495 2040 0.625 0.572 0.533 0.642 0.591 0.555 2045 0.651 0.613 0.588 0.671 0.636 0.614 2050 0.662 0.635 0.628 0.683 0.660 0.659 Table A36
Summary of Alternative Population Projections Indonesia Medium Fertility Scenario
Population (1000s) Percentage Year 0-14 15-64 65+ 0-14 15-64 65+ Standard Mortality Scenario 1995 65,041 123,892 8,532 32.9 62.7 4.3 2000 64,926 137,181 10,001 30.6 64.7 4.7 2005 63,923 149,532 12,021 28.4 66.3 5.3 2010 62,753 161,275 13,984 26.4 67.8 5.9 2015 61,774 172,825 15,785 24.7 69.0 6.3 2020 61,918 181,603 18,770 23.6 69.2 7.2 2025 62,562 187,802 23,079 22.9 68.7 8.4 2030 62,538 193,034 27,950 22.1 68.1 9.9 2035 62,182 196,431 33,814 21.3 67.2 11.6 2040 61,872 197,838 40,411 20.6 65.9 13.5 2045 61,870 198,332 46,388 20.2 64.7 15.1 2050 62,050 198,307 51,495 19.9 63.6 16.5 High Mortality Scenario 1995 65,041 123,892 8,532 32.9 62.7 4.3 2000 64,926 137,181 10,001 30.6 64.7 4.7 2005 63,744 149,301 11,950 28.3 66.4 5.3 2010 62,291 160,551 13,743 26.3 67.9 5.8 2015 61,017 171,492 15,302 24.6 69.2 6.2 2020 60,938 179,518 17,945 23.6 69.5 6.9 2025 61,355 184,821 21,728 22.9 69.0 8.1 2030 61,139 189,074 25,809 22.1 68.5 9.4 2035 60,565 191,517 30,569 21.4 67.8 10.8 2040 60,024 192,119 35,731 20.9 66.7 12.4 2045 59,783 191,882 40,141 20.5 65.8 13.8 2050 59,764 191,192 43,598 20.3 64.9 14.8 Low Mortality Scenario 1995 65,041 123,892 8,532 32.9 62.7 4.3 2000 64,926 137,181 10,001 30.6 64.7 4.7 2005 64,207 149,993 12,173 28.4 66.3 5.4 2010 63,367 162,318 14,419 26.4 67.6 6.0 2015 62,740 174,672 16,743 24.7 68.7 6.6 2020 63,072 184,412 20,584 23.5 68.8 7.7 2025 63,776 191,542 26,041 22.7 68.1 9.3 Table A37
Summary of Alternative Population Projections Indonesia High Fertility Scenario
Population (1000s) Percentage Year 0-14 15-64 65+ 0-14 15-64 65+ Standard Mortality Scenario 1995 65,041 123,892 8,532 32.9 62.7 4.3 2000 66,425 137,181 10,001 31.1 64.2 4.7 2005 68,805 149,532 12,021 29.9 64.9 5.2 2010 72,209 161,275 13,984 29.2 65.2 5.7 2015 74,639 174,302 15,785 28.2 65.8 6.0 2020 76,920 186,431 18,770 27.3 66.1 6.7 2025 79,601 197,169 23,079 26.5 65.8 7.7 2030 82,684 207,255 27,950 26.0 65.2 8.8 2035 86,161 216,106 33,814 25.6 64.3 10.1 2040 89,726 224,048 40,411 25.3 63.3 11.4 2045 93,303 232,449 46,388 25.1 62.5 12.5 2050 96,961 241,635 51,495 24.9 61.9 13.2 High Mortality Scenario 1995 65,041 123,892 8,532 32.9 62.7 4.3 2000 66,425 137,181 10,001 31.1 64.2 4.7 2005 68,600 149,301 11,950 29.8 65.0 5.2 2010 71,652 160,551 13,743 29.1 65.3 5.6 2015 73,697 172,948 15,302 28.1 66.0 5.8 2020 75,682 184,254 17,945 27.2 66.3 6.5 2025 78,044 193,973 21,728 26.6 66.0 7.4 2030 80,788 202,912 25,809 26.1 65.6 8.3 2035 83,818 210,595 30,569 25.8 64.8 9.4 2040 86,879 217,440 35,731 25.5 63.9 10.5 2045 89,951 224,701 40,141 25.4 63.3 11.3 2050 93,174 232,662 43,598 25.2 63.0 11.8 Low Mortality Scenario 1995 65,041 123,892 8,532 32.9 62.7 4.3 2000 66,425 137,181 10,001 31.1 64.2 4.7 2005 69,133 149,993 12,173 29.9 64.8 5.3 2010 72,948 162,336 14,419 29.2 65.0 5.8 2015 75,821 176,221 16,743 28.2 65.6 6.2 2020 78,394 189,420 20,584 27.2 65.7 7.1 2025 81,241 201,272 26,041 26.3 65.2 8.4 Table A38
Summary of Alternative Population Projections Indonesia Low Fertility Scenario
Population (1000s) Percentage Year 0-14 15-64 65+ 0-14 15-64 65+ Standard Mortality Scenario 1995 65,041 123,892 8,532 32.9 62.7 4.3 2000 63,396 137,181 10,001 30.1 65.1 4.7 2005 58,793 149,532 12,021 26.7 67.9 5.5 2010 52,790 161,275 13,984 23.1 70.7 6.1 2015 48,315 171,319 15,785 20.5 72.8 6.7 2020 46,641 176,532 18,770 19.3 73.0 7.8 2025 45,954 177,932 23,079 18.6 72.0 9.3 2030 43,847 178,195 27,950 17.5 71.3 11.2 2035 40,967 176,243 33,814 16.3 70.2 13.5 2040 38,257 171,560 40,411 15.3 68.6 16.1 2045 36,260 165,051 46,388 14.6 66.6 18.7 2050 34,733 157,226 51,495 14.3 64.6 21.2 High Mortality Scenario 1995 65,041 123,892 8,532 32.9 62.7 4.3 2000 63,396 137,181 10,001 30.1 65.1 4.7 2005 58,644 149,301 11,950 26.7 67.9 5.4 2010 52,431 160,551 13,743 23.1 70.8 6.1 2015 47,756 170,006 15,302 20.5 72.9 6.6 2020 45,931 174,546 17,945 19.3 73.2 7.5 2025 45,089 175,185 21,728 18.6 72.4 9.0 2030 42,900 174,647 25,809 17.6 71.8 10.6 2035 39,977 171,960 30,569 16.5 70.9 12.6 2040 37,240 166,756 35,731 15.5 69.6 14.9 2045 35,193 159,895 40,141 15.0 68.0 17.1 2050 33,617 151,908 43,598 14.7 66.3 19.0 Low Mortality Scenario 1995 65,041 123,892 8,532 32.9 62.7 4.3 2000 63,396 137,181 10,001 30.1 65.1 4.7 2005 59,027 149,993 12,173 26.7 67.8 5.5 2010 53,259 162,294 14,419 23.2 70.6 6.3 2015 49,049 173,061 16,743 20.5 72.5 7.0 2020 47,459 179,087 20,584 19.2 72.5 8.3 2025 46,728 181,180 26,041 18.4 71.3 10.3 Table A39
Summary of Aging, Alternative Mortality Scenarios Indonesia
Percentage 65 and older Old Age Dependency Ratio Year Medium High Low Medium High Low Medium Fertility Scenario 1995 4.3 4.3 4.3 0.069 0.069 0.069 2000 4.7 4.7 4.7 0.073 0.073 0.073 2005 5.3 5.3 5.4 0.080 0.080 0.081 2010 5.9 5.8 6.0 0.087 0.086 0.089 2015 6.3 6.2 6.6 0.091 0.089 0.096 2020 7.2 6.9 7.7 0.103 0.100 0.112 2025 8.4 8.1 9.3 0.123 0.118 0.136 2030 9.9 9.4 0.145 0.137 2035 11.6 10.8 0.172 0.160 2040 13.5 12.4 0.204 0.186 2045 15.1 13.8 0.234 0.209 2050 16.5 14.8 0.260 0.228 High Fertility Scenario 1995 4.3 4.3 4.3 0.069 0.069 0.069 2000 4.7 4.7 4.7 0.073 0.073 0.073 2005 5.2 5.2 5.3 0.080 0.080 0.081 2010 5.7 5.6 5.8 0.087 0.086 0.089 2015 6.0 5.8 6.2 0.091 0.088 0.095 2020 6.7 6.5 7.1 0.101 0.097 0.109 2025 7.7 7.4 8.4 0.117 0.112 0.129 2030 8.8 8.3 0.135 0.127 2035 10.1 9.4 0.156 0.145 2040 11.4 10.5 0.180 0.164 2045 12.5 11.3 0.200 0.179 2050 13.2 11.8 0.213 0.187 Low Fertility Scenario 1995 4.3 4.3 4.3 0.069 0.069 0.069 2000 4.7 4.7 4.7 0.073 0.073 0.073 2005 5.5 5.4 5.5 0.080 0.080 0.081 2010 6.1 6.1 6.3 0.087 0.086 0.089 2015 6.7 6.6 7.0 0.092 0.090 0.097 2020 7.8 7.5 8.3 0.106 0.103 0.115 2025 9.3 9.0 10.3 0.130 0.124 0.144 2030 11.2 10.6 0.157 0.148 2035 13.5 12.6 0.192 0.178 2040 16.1 14.9 0.236 0.214 2045 18.7 17.1 0.281 0.251 2050 21.2 19.0 0.328 0.287 Table A40
Total Dependency Ratio Indonesia
Fertility: High Med Low High Med Low High Med Low Mortality: High High High Med Med Med Low Low Low
1995 0.594 0.594 0.594 0.594 0.594 0.594 0.594 0.594 0.594 2000 0.557 0.546 0.535 0.557 0.546 0.535 0.557 0.546 0.535 2005 0.540 0.507 0.473 0.541 0.508 0.474 0.542 0.509 0.475 2010 0.532 0.474 0.412 0.534 0.476 0.414 0.538 0.479 0.417 2015 0.515 0.445 0.371 0.519 0.449 0.374 0.530 0.455 0.380 2020 0.508 0.439 0.366 0.513 0.444 0.371 0.537 0.454 0.380 2025 0.514 0.450 0.381 0.521 0.456 0.388 0.560 0.469 0.402 2030 0.525 0.460 0.393 0.534 0.469 0.403 2035 0.543 0.476 0.410 0.555 0.489 0.424 2040 0.564 0.498 0.438 0.581 0.517 0.459 2045 0.579 0.521 0.471 0.601 0.546 0.501 2050 0.588 0.541 0.508 0.614 0.573 0.548 Table A41
Summary of Alternative Population Projections India Medium Fertility Scenario
Population (1000s) Percentage Year 0-14 15-64 65+ 0-14 15-64 65+ Standard Mortality Scenario 1995 330,496 560,252 42,915 35.4 60.0 4.6 2000 337,607 625,589 50,461 33.3 61.7 5.0 2005 335,390 693,626 58,437 30.8 63.8 5.4 2010 323,604 761,473 67,077 28.1 66.1 5.8 2015 311,090 822,753 77,808 25.7 67.9 6.4 2020 305,228 874,247 92,672 24.0 68.7 7.3 2025 306,458 912,057 111,908 23.0 68.6 8.4 2030 308,436 940,539 133,711 22.3 68.0 9.7 2035 304,976 965,757 157,236 21.4 67.6 11.0 2040 300,506 985,477 181,062 20.5 67.2 12.3 2045 298,702 998,666 203,537 19.9 66.5 13.6 2050 299,118 998,468 231,164 19.6 65.3 15.1 High Mortality Scenario 1995 330,496 560,252 42,915 35.4 60.0 4.6 2000 337,607 625,589 50,461 33.3 61.7 5.0 2005 334,428 693,321 58,028 30.8 63.9 5.3 2010 321,055 760,149 65,784 28.0 66.3 5.7 2015 307,020 819,321 75,162 25.6 68.2 6.3 2020 300,385 866,643 87,986 23.9 69.1 7.0 2025 300,568 898,857 104,390 23.1 68.9 8.0 2030 301,022 921,278 122,788 22.4 68.5 9.1 2035 295,590 940,605 142,330 21.4 68.2 10.3 2040 289,336 954,710 161,818 20.6 67.9 11.5 2045 286,110 962,730 179,548 20.0 67.4 12.6 2050 285,611 958,505 201,736 19.8 66.3 14.0 Low Mortality Scenario 1995 330,496 560,252 42,915 35.4 60.0 4.6 2000 337,607 625,589 50,461 33.3 61.7 5.0 2005 337,202 694,468 59,283 30.9 63.7 5.4 2010 326,994 765,268 69,694 28.1 65.9 6.0 2015 316,244 831,517 82,962 25.7 67.6 6.7 2020 312,031 888,315 100,818 24.0 68.3 7.7 2025 314,390 930,604 123,700 23.0 68.0 9.0 Table A42
Summary of Alternative Population Projections India High Fertility Scenario
Population (1000s) Percentage Year 0-14 15-64 65+ 0-14 15-64 65+ Standard Mortality Scenario 1995 330,496 560,252 42,915 35.4 60.0 4.6 2000 342,770 625,589 50,461 33.6 61.4 5.0 2005 351,635 693,626 58,437 31.9 62.8 5.3 2010 358,059 761,473 67,077 30.2 64.2 5.7 2015 363,944 827,759 77,808 28.7 65.2 6.1 2020 373,905 890,137 92,672 27.6 65.6 6.8 2025 389,000 945,914 111,908 26.9 65.4 7.7 2030 405,857 997,544 133,711 26.4 64.9 8.7 2035 420,259 1,049,267 157,236 25.8 64.5 9.7 2040 434,472 1,100,553 181,062 25.3 64.1 10.6 2045 451,236 1,151,373 203,537 25.0 63.7 11.3 2050 470,501 1,194,977 231,164 24.8 63.0 12.2 High Mortality Scenario 1995 330,496 560,252 42,915 35.4 60.0 4.6 2000 342,770 625,589 50,461 33.6 61.4 5.0 2005 350,605 693,321 58,028 31.8 62.9 5.3 2010 355,194 760,149 65,784 30.1 64.4 5.6 2015 359,153 824,288 75,162 28.5 65.5 6.0 2020 367,952 882,317 87,986 27.5 65.9 6.6 2025 381,478 932,089 104,390 26.9 65.7 7.4 2030 396,007 977,033 122,788 26.5 65.3 8.2 2035 407,153 1,022,012 142,330 25.9 65.0 9.1 2040 418,092 1,066,409 161,818 25.4 64.8 9.8 2045 431,955 1,110,240 179,548 25.1 64.5 10.4 2050 448,948 1,147,293 201,736 25.0 63.8 11.2 Low Mortality Scenario 1995 330,496 560,252 42,915 35.4 60.0 4.6 2000 342,770 625,589 50,461 33.6 61.4 5.0 2005 353,583 694,468 59,283 31.9 62.7 5.4 2010 361,871 765,283 69,694 30.2 63.9 5.8 2015 370,025 836,638 82,962 28.7 64.9 6.4 2020 382,309 904,608 100,818 27.5 65.2 7.3 2025 399,176 965,399 123,700 26.8 64.9 8.3 Table A43
Summary of Alternative Population Projections India Low Fertility Scenario
Population (1000s) Percentage Year 0-14 15-64 65+ 0-14 15-64 65+ Standard Mortality Scenario 1995 330,496 560,252 42,915 35.4 60.0 4.6 2000 332,415 625,589 50,461 33.0 62.0 5.0 2005 319,001 693,626 58,437 29.8 64.8 5.5 2010 288,470 761,473 67,077 25.8 68.2 6.0 2015 256,969 817,719 77,808 22.3 71.0 6.8 2020 235,878 858,217 92,672 19.9 72.3 7.8 2025 226,205 877,534 111,908 18.6 72.2 9.2 2030 218,863 882,258 133,711 17.7 71.4 10.8 2035 204,646 881,439 157,236 16.5 70.9 12.6 2040 189,255 872,004 181,062 15.2 70.2 14.6 2045 177,201 852,474 203,537 14.4 69.1 16.5 2050 168,366 816,003 231,164 13.9 67.1 19.0 High Mortality Scenario 1995 330,496 560,252 42,915 35.4 60.0 4.6 2000 332,415 625,589 50,461 33.0 62.0 5.0 2005 318,108 693,321 58,028 29.7 64.8 5.4 2010 286,245 760,149 65,784 25.7 68.3 5.9 2015 253,641 814,327 75,162 22.2 71.2 6.6 2020 232,161 850,831 87,986 19.8 72.7 7.5 2025 221,902 864,976 104,390 18.6 72.6 8.8 2030 213,672 864,284 122,788 17.8 72.0 10.2 2035 198,462 858,426 142,330 16.5 71.6 11.9 2040 182,370 844,583 161,818 15.3 71.0 13.6 2045 169,905 821,516 179,548 14.5 70.2 15.3 2050 160,970 783,180 201,736 14.0 68.3 17.6 Low Mortality Scenario 1995 330,496 560,252 42,915 35.4 60.0 4.6 2000 332,415 625,589 50,461 33.0 62.0 5.0 2005 320,673 694,468 59,283 29.8 64.6 5.5 2010 291,418 765,251 69,694 25.9 67.9 6.2 2015 261,161 826,341 82,962 22.3 70.6 7.1 2020 241,056 871,796 100,818 19.9 71.8 8.3 2025 231,939 894,972 123,700 18.5 71.6 9.9 Table A44
Summary of Aging, Alternative Mortality Scenarios India
Percentage 65 and older Old Age Dependency Ratio Year Medium High Low Medium High Low Medium Fertility Scenario 1995 4.6 4.6 4.6 0.077 0.077 0.077 2000 5.0 5.0 5.0 0.081 0.081 0.081 2005 5.4 5.3 5.4 0.084 0.084 0.085 2010 5.8 5.7 6.0 0.088 0.087 0.091 2015 6.4 6.3 6.7 0.095 0.092 0.100 2020 7.3 7.0 7.7 0.106 0.102 0.113 2025 8.4 8.0 9.0 0.123 0.116 0.133 2030 9.7 9.1 0.142 0.133 2035 11.0 10.3 0.163 0.151 2040 12.3 11.5 0.184 0.169 2045 13.6 12.6 0.204 0.186 2050 15.1 14.0 0.232 0.210 High Fertility Scenario 1995 4.6 4.6 4.6 0.077 0.077 0.077 2000 5.0 5.0 5.0 0.081 0.081 0.081 2005 5.3 5.3 5.4 0.084 0.084 0.085 2010 5.7 5.6 5.8 0.088 0.087 0.091 2015 6.1 6.0 6.4 0.094 0.091 0.099 2020 6.8 6.6 7.3 0.104 0.100 0.111 2025 7.7 7.4 8.3 0.118 0.112 0.128 2030 8.7 8.2 0.134 0.126 2035 9.7 9.1 0.150 0.139 2040 10.6 9.8 0.165 0.152 2045 11.3 10.4 0.177 0.162 2050 12.2 11.2 0.193 0.176 Low Fertility Scenario 1995 4.6 4.6 4.6 0.077 0.077 0.077 2000 5.0 5.0 5.0 0.081 0.081 0.081 2005 5.5 5.4 5.5 0.084 0.084 0.085 2010 6.0 5.9 6.2 0.088 0.087 0.091 2015 6.8 6.6 7.1 0.095 0.092 0.100 2020 7.8 7.5 8.3 0.108 0.103 0.116 2025 9.2 8.8 9.9 0.128 0.121 0.138 2030 10.8 10.2 0.152 0.142 2035 12.6 11.9 0.178 0.166 2040 14.6 13.6 0.208 0.192 2045 16.5 15.3 0.239 0.219 2050 19.0 17.6 0.283 0.258 Table A45
Total Dependency Ratio India
Fertility: High Med Low High Med Low High Med Low Mortality: High High High Med Med Med Low Low Low
1995 0.667 0.667 0.667 0.667 0.667 0.667 0.667 0.667 0.667 2000 0.629 0.620 0.612 0.629 0.620 0.612 0.629 0.620 0.612 2005 0.589 0.566 0.543 0.591 0.568 0.544 0.595 0.571 0.547 2010 0.554 0.509 0.463 0.558 0.513 0.467 0.564 0.518 0.472 2015 0.527 0.466 0.404 0.534 0.473 0.409 0.545 0.480 0.416 2020 0.517 0.448 0.376 0.524 0.455 0.383 0.544 0.465 0.392 2025 0.521 0.451 0.377 0.530 0.459 0.385 0.562 0.471 0.397 2030 0.531 0.460 0.389 0.541 0.470 0.400 2035 0.538 0.466 0.397 0.550 0.479 0.411 2040 0.544 0.473 0.408 0.559 0.489 0.425 2045 0.551 0.484 0.425 0.569 0.503 0.447 2050 0.567 0.508 0.463 0.587 0.531 0.490 Table A46
Summary of Alternative Population Projections Bangladesh Medium Fertility Scenario
Population (1000s) Percentage Year 0-14 15-64 65+ 0-14 15-64 65+ Standard Mortality Scenario 1995 47,458 67,418 3,740 40.0 56.8 3.2 2000 45,358 79,648 4,149 35.1 61.7 3.2 2005 45,009 90,743 4,813 32.0 64.6 3.4 2010 47,303 98,791 5,706 31.2 65.1 3.8 2015 46,938 107,695 6,908 29.1 66.7 4.3 2020 44,787 116,854 8,554 26.3 68.7 5.0 2025 43,050 125,207 10,497 24.1 70.0 5.9 2030 42,967 131,635 12,549 23.0 70.3 6.7 2035 43,799 136,217 15,010 22.5 69.8 7.7 2040 44,080 138,693 19,111 21.8 68.7 9.5 2045 43,697 140,156 23,822 21.0 67.5 11.5 2050 43,167 139,541 29,794 20.3 65.7 14.0 High Mortality Scenario 1995 47,458 67,418 3,740 40.0 56.8 3.2 2000 45,358 79,648 4,149 35.1 61.7 3.2 2005 44,817 90,564 4,781 32.0 64.6 3.4 2010 46,721 98,194 5,601 31.0 65.2 3.7 2015 45,921 106,478 6,683 28.9 66.9 4.2 2020 43,436 114,749 8,128 26.1 69.0 4.9 2025 41,420 121,986 9,777 23.9 70.4 5.6 2030 41,054 127,190 11,450 22.8 70.8 6.4 2035 41,541 130,474 13,419 22.4 70.4 7.2 2040 41,480 131,770 16,796 21.8 69.3 8.8 2045 40,790 132,084 20,533 21.1 68.3 10.6 2050 40,033 130,553 25,125 20.5 66.7 12.8 Low Mortality Scenario 1995 47,458 67,418 3,740 40.0 56.8 3.2 2000 45,358 79,648 4,149 35.1 61.7 3.2 2005 45,381 91,121 4,877 32.1 64.5 3.4 2010 48,259 99,779 5,909 31.3 64.8 3.8 2015 48,424 109,509 7,352 29.3 66.3 4.4 2020 46,505 119,794 9,361 26.5 68.2 5.3 2025 44,781 129,442 11,876 24.1 69.6 6.4 Table A47
Summary of Alternative Population Projections Bangladesh High Fertility Scenario
Population (1000s) Percentage Year 0-14 15-64 65+ 0-14 15-64 65+ Standard Mortality Scenario 1995 47,458 67,418 3,740 40.0 56.8 3.2 2000 46,347 79,648 4,149 35.6 61.2 3.2 2005 47,631 90,743 4,813 33.3 63.4 3.4 2010 52,187 98,791 5,706 33.3 63.1 3.6 2015 53,857 108,651 6,908 31.8 64.1 4.1 2020 53,809 119,414 8,554 29.6 65.7 4.7 2025 54,241 130,001 10,497 27.9 66.8 5.4 2030 56,371 139,390 12,549 27.1 66.9 6.0 2035 59,488 147,648 15,010 26.8 66.5 6.8 2040 62,176 154,495 19,111 26.4 65.5 8.1 2045 64,320 161,093 23,822 25.8 64.6 9.6 2050 66,469 166,384 29,794 25.3 63.4 11.3 High Mortality Scenario 1995 47,458 67,418 3,740 40.0 56.8 3.2 2000 46,347 79,648 4,149 35.6 61.2 3.2 2005 47,420 90,564 4,781 33.2 63.4 3.3 2010 51,533 98,194 5,601 33.2 63.2 3.6 2015 52,678 107,416 6,683 31.6 64.4 4.0 2020 52,168 117,243 8,128 29.4 66.0 4.6 2025 52,164 126,618 9,777 27.7 67.2 5.2 2030 53,821 134,617 11,450 26.9 67.3 5.7 2035 56,359 141,332 13,419 26.7 66.9 6.4 2040 58,423 146,668 16,796 26.3 66.1 7.6 2045 59,936 151,673 20,533 25.8 65.3 8.8 2050 61,523 155,471 25,125 25.4 64.2 10.4 Low Mortality Scenario 1995 47,458 67,418 3,740 40.0 56.8 3.2 2000 46,347 79,648 4,149 35.6 61.2 3.2 2005 48,037 91,121 4,877 33.4 63.3 3.4 2010 53,263 99,789 5,909 33.5 62.8 3.7 2015 55,587 110,505 7,352 32.0 63.7 4.2 2020 55,899 122,482 9,361 29.8 65.2 5.0 2025 56,462 134,521 11,876 27.8 66.3 5.9 Table A48
Summary of Alternative Population Projections Bangladesh Low Fertility Scenario
Population (1000s) Percentage Year 0-14 15-64 65+ 0-14 15-64 65+ Standard Mortality Scenario 1995 47,458 67,418 3,740 40.0 56.8 3.2 2000 44,270 79,648 4,149 34.6 62.2 3.2 2005 42,236 90,743 4,813 30.7 65.9 3.5 2010 42,245 98,791 5,706 28.8 67.3 3.9 2015 39,948 106,644 6,908 26.0 69.5 4.5 2020 35,903 114,146 8,554 22.6 72.0 5.4 2025 32,434 120,243 10,497 19.9 73.7 6.4 2030 30,773 123,717 12,549 18.4 74.1 7.5 2035 30,087 124,776 15,010 17.7 73.5 8.8 2040 28,857 123,295 19,111 16.9 72.0 11.2 2045 27,116 120,261 23,822 15.8 70.2 13.9 2050 25,325 114,658 29,794 14.9 67.5 17.5 High Mortality Scenario 1995 47,458 67,418 3,740 40.0 56.8 3.2 2000 44,270 79,648 4,149 34.6 62.2 3.2 2005 42,064 90,564 4,781 30.6 65.9 3.5 2010 41,739 98,194 5,601 28.7 67.5 3.8 2015 39,097 105,446 6,683 25.9 69.7 4.4 2020 34,838 112,111 8,128 22.5 72.3 5.2 2025 31,232 117,191 9,777 19.7 74.1 6.2 2030 29,438 119,610 11,450 18.3 74.5 7.1 2035 28,585 119,614 13,419 17.7 74.0 8.3 2040 27,218 117,265 16,796 16.9 72.7 10.4 2045 25,387 113,484 20,533 15.9 71.2 12.9 2050 23,572 107,472 25,125 15.1 68.8 16.1 Low Mortality Scenario 1995 47,458 67,418 3,740 40.0 56.8 3.2 2000 44,270 79,648 4,149 34.6 62.2 3.2 2005 42,572 91,121 4,877 30.7 65.8 3.5 2010 43,076 99,768 5,909 29.0 67.1 4.0 2015 41,178 108,409 7,352 26.2 69.1 4.7 2020 37,242 116,928 9,361 22.8 71.5 5.7 2025 33,689 124,130 11,876 19.9 73.1 7.0 Table A49
Summary of Aging, Alternative Mortality Scenarios Bangladesh
Percentage 65 and older Old Age Dependency Ratio Year Medium High Low Medium High Low Medium Fertility Scenario 1995 3.2 3.2 3.2 0.055 0.055 0.055 2000 3.2 3.2 3.2 0.052 0.052 0.052 2005 3.4 3.4 3.4 0.053 0.053 0.054 2010 3.8 3.7 3.8 0.058 0.057 0.059 2015 4.3 4.2 4.4 0.064 0.063 0.067 2020 5.0 4.9 5.3 0.073 0.071 0.078 2025 5.9 5.6 6.4 0.084 0.080 0.092 2030 6.7 6.4 0.095 0.090 2035 7.7 7.2 0.110 0.103 2040 9.5 8.8 0.138 0.127 2045 11.5 10.6 0.170 0.155 2050 14.0 12.8 0.214 0.192 High Fertility Scenario 1995 3.2 3.2 3.2 0.055 0.055 0.055 2000 3.2 3.2 3.2 0.052 0.052 0.052 2005 3.4 3.3 3.4 0.053 0.053 0.054 2010 3.6 3.6 3.7 0.058 0.057 0.059 2015 4.1 4.0 4.2 0.064 0.062 0.067 2020 4.7 4.6 5.0 0.072 0.069 0.076 2025 5.4 5.2 5.9 0.081 0.077 0.088 2030 6.0 5.7 0.090 0.085 2035 6.8 6.4 0.102 0.095 2040 8.1 7.6 0.124 0.115 2045 9.6 8.8 0.148 0.135 2050 11.3 10.4 0.179 0.162 Low Fertility Scenario 1995 3.2 3.2 3.2 0.055 0.055 0.055 2000 3.2 3.2 3.2 0.052 0.052 0.052 2005 3.5 3.5 3.5 0.053 0.053 0.054 2010 3.9 3.8 4.0 0.058 0.057 0.059 2015 4.5 4.4 4.7 0.065 0.063 0.068 2020 5.4 5.2 5.7 0.075 0.073 0.080 2025 6.4 6.2 7.0 0.087 0.083 0.096 2030 7.5 7.1 0.101 0.096 2035 8.8 8.3 0.120 0.112 2040 11.2 10.4 0.155 0.143 2045 13.9 12.9 0.198 0.181 2050 17.5 16.1 0.260 0.234 Table A50
Total Dependency Ratio Bangladesh
Fertility: High Med Low High Med Low High Med Low Mortality: High High High Med Med Med Low Low Low
1995 0.759 0.759 0.759 0.759 0.759 0.759 0.759 0.759 0.759 2000 0.634 0.622 0.608 0.634 0.622 0.608 0.634 0.622 0.608 2005 0.576 0.548 0.517 0.578 0.549 0.518 0.581 0.552 0.521 2010 0.582 0.533 0.482 0.586 0.537 0.485 0.593 0.543 0.491 2015 0.553 0.494 0.434 0.559 0.500 0.439 0.575 0.509 0.448 2020 0.514 0.449 0.383 0.522 0.456 0.389 0.545 0.466 0.399 2025 0.489 0.420 0.350 0.498 0.428 0.357 0.528 0.438 0.367 2030 0.485 0.413 0.342 0.494 0.422 0.350 2035 0.494 0.421 0.351 0.505 0.432 0.361 2040 0.513 0.442 0.375 0.526 0.456 0.389 2045 0.531 0.464 0.405 0.547 0.482 0.424 2050 0.557 0.499 0.453 0.579 0.523 0.481 Table A51
Summary of Alternative Population Projections Egypt Medium Fertility Scenario
Population (1000s) Percentage Year 0-14 15-64 65+ 0-14 15-64 65+ Standard Mortality Scenario 1995 23,830 36,005 2,447 38.3 57.8 3.9 2000 24,195 41,460 2,814 35.3 60.6 4.1 2005 23,903 47,441 3,191 32.1 63.6 4.3 2010 23,611 52,840 3,613 29.5 66.0 4.5 2015 22,952 57,842 4,431 26.9 67.9 5.2 2020 22,696 62,028 5,769 25.1 68.5 6.4 2025 22,976 65,309 7,333 24.0 68.3 7.7 2030 23,401 68,186 8,788 23.3 67.9 8.8 2035 23,374 70,899 10,344 22.3 67.8 9.9 2040 23,171 73,382 11,821 21.4 67.7 10.9 2045 23,055 74,878 13,857 20.6 67.0 12.4 2050 23,125 75,115 16,609 20.1 65.4 14.5 High Mortality Scenario 1995 23,830 36,005 2,447 38.3 57.8 3.9 2000 24,195 41,460 2,814 35.3 60.6 4.1 2005 23,830 47,369 3,166 32.0 63.7 4.3 2010 23,439 52,669 3,531 29.4 66.1 4.4 2015 22,679 57,526 4,266 26.8 68.1 5.1 2020 22,360 61,496 5,477 25.0 68.8 6.1 2025 22,592 64,528 6,831 24.0 68.7 7.3 2030 22,965 67,135 8,005 23.4 68.4 8.2 2035 22,882 69,576 9,204 22.5 68.4 9.1 2040 22,611 71,786 10,291 21.6 68.6 9.8 2045 22,428 73,033 11,860 20.9 68.1 11.1 2050 22,448 73,091 14,049 20.5 66.7 12.8 Low Mortality Scenario 1995 23,830 36,005 2,447 38.3 57.8 3.9 2000 24,195 41,460 2,814 35.3 60.6 4.1 2005 24,003 47,529 3,245 32.1 63.6 4.3 2010 23,832 53,073 3,756 29.5 65.8 4.7 2015 23,278 58,262 4,755 27.0 67.5 5.5 2020 23,043 62,690 6,369 25.0 68.1 6.9 2025 23,310 66,202 8,307 23.8 67.7 8.5 Table A52
Summary of Alternative Population Projections Egypt High Fertility Scenario
Population (1000s) Percentage Year 0-14 15-64 65+ 0-14 15-64 65+ Standard Mortality Scenario 1995 23,830 36,005 2,447 38.3 57.8 3.9 2000 24,643 41,460 2,814 35.8 60.2 4.1 2005 25,183 47,441 3,191 33.2 62.6 4.2 2010 26,229 52,840 3,613 31.7 63.9 4.4 2015 26,851 58,284 4,431 30.0 65.1 4.9 2020 27,734 63,297 5,769 28.7 65.4 6.0 2025 29,002 67,909 7,333 27.8 65.1 7.0 2030 30,467 72,502 8,788 27.3 64.9 7.9 2035 31,710 77,175 10,344 26.6 64.7 8.7 2040 32,889 81,967 11,821 26.0 64.7 9.3 2045 34,169 86,205 13,857 25.5 64.2 10.3 2050 35,630 89,649 16,609 25.1 63.2 11.7 High Mortality Scenario 1995 23,830 36,005 2,447 38.3 57.8 3.9 2000 24,643 41,460 2,814 35.8 60.2 4.1 2005 25,103 47,369 3,166 33.2 62.6 4.2 2010 26,032 52,669 3,531 31.7 64.0 4.3 2015 26,526 57,965 4,266 29.9 65.3 4.8 2020 27,319 62,751 5,477 28.6 65.7 5.7 2025 28,512 67,086 6,831 27.8 65.5 6.7 2030 29,890 71,368 8,005 27.4 65.3 7.3 2035 31,023 75,712 9,204 26.8 65.3 7.9 2040 32,063 80,157 10,291 26.2 65.4 8.4 2045 33,197 84,046 11,860 25.7 65.1 9.2 2050 34,537 87,174 14,049 25.4 64.2 10.3 Low Mortality Scenario 1995 23,830 36,005 2,447 38.3 57.8 3.9 2000 24,643 41,460 2,814 35.8 60.2 4.1 2005 25,293 47,529 3,245 33.3 62.5 4.3 2010 26,482 53,075 3,756 31.8 63.7 4.5 2015 27,238 58,716 4,755 30.0 64.7 5.2 2020 28,166 64,000 6,369 28.6 65.0 6.5 2025 29,441 68,889 8,307 27.6 64.6 7.8 Table A53
Summary of Alternative Population Projections Egypt Low Fertility Scenario
Population (1000s) Percentage Year 0-14 15-64 65+ 0-14 15-64 65+ Standard Mortality Scenario 1995 23,830 36,005 2,447 38.3 57.8 3.9 2000 23,745 41,460 2,814 34.9 61.0 4.1 2005 22,608 47,441 3,191 30.9 64.8 4.4 2010 20,937 52,840 3,613 27.1 68.3 4.7 2015 18,944 57,397 4,431 23.5 71.1 5.5 2020 17,579 60,744 5,769 20.9 72.2 6.9 2025 17,060 62,652 7,333 19.6 72.0 8.4 2030 16,812 63,758 8,788 18.8 71.4 9.8 2035 15,996 64,531 10,344 17.6 71.0 11.4 2040 14,954 64,851 11,821 16.3 70.8 12.9 2045 14,028 63,916 13,857 15.3 69.6 15.1 2050 13,354 61,444 16,609 14.6 67.2 18.2 High Mortality Scenario 1995 23,830 36,005 2,447 38.3 57.8 3.9 2000 23,745 41,460 2,814 34.9 61.0 4.1 2005 22,542 47,369 3,166 30.8 64.8 4.3 2010 20,790 52,669 3,531 27.0 68.4 4.6 2015 18,725 57,083 4,266 23.4 71.3 5.3 2020 17,325 60,227 5,477 20.9 72.5 6.6 2025 16,781 61,913 6,831 19.6 72.4 8.0 2030 16,507 62,794 8,005 18.9 71.9 9.2 2035 15,673 63,352 9,204 17.8 71.8 10.4 2040 14,614 63,470 10,291 16.5 71.8 11.6 2045 13,676 62,379 11,860 15.6 71.0 13.5 2050 12,999 59,849 14,049 15.0 68.9 16.2 Low Mortality Scenario 1995 23,830 36,005 2,447 38.3 57.8 3.9 2000 23,745 41,460 2,814 34.9 61.0 4.1 2005 22,697 47,529 3,245 30.9 64.7 4.4 2010 21,123 53,070 3,756 27.1 68.1 4.8 2015 19,206 57,801 4,755 23.5 70.7 5.8 2020 17,840 61,353 6,369 20.9 71.7 7.4 2025 17,289 63,431 8,307 19.4 71.2 9.3 Table A54
Summary of Aging, Alternative Mortality Scenarios Egypt
Percentage 65 and older Old Age Dependency Ratio Year Medium High Low Medium High Low Medium Fertility Scenario 1995 3.9 3.9 3.9 0.068 0.068 0.068 2000 4.1 4.1 4.1 0.068 0.068 0.068 2005 4.3 4.3 4.3 0.067 0.067 0.068 2010 4.5 4.4 4.7 0.068 0.067 0.071 2015 5.2 5.1 5.5 0.077 0.074 0.082 2020 6.4 6.1 6.9 0.093 0.089 0.102 2025 7.7 7.3 8.5 0.112 0.106 0.125 2030 8.8 8.2 0.129 0.119 2035 9.9 9.1 0.146 0.132 2040 10.9 9.8 0.161 0.143 2045 12.4 11.1 0.185 0.162 2050 14.5 12.8 0.221 0.192 High Fertility Scenario 1995 3.9 3.9 3.9 0.068 0.068 0.068 2000 4.1 4.1 4.1 0.068 0.068 0.068 2005 4.2 4.2 4.3 0.067 0.067 0.068 2010 4.4 4.3 4.5 0.068 0.067 0.071 2015 4.9 4.8 5.2 0.076 0.074 0.081 2020 6.0 5.7 6.5 0.091 0.087 0.100 2025 7.0 6.7 7.8 0.108 0.102 0.121 2030 7.9 7.3 0.121 0.112 2035 8.7 7.9 0.134 0.122 2040 9.3 8.4 0.144 0.128 2045 10.3 9.2 0.161 0.141 2050 11.7 10.3 0.185 0.161 Low Fertility Scenario 1995 3.9 3.9 3.9 0.068 0.068 0.068 2000 4.1 4.1 4.1 0.068 0.068 0.068 2005 4.4 4.3 4.4 0.067 0.067 0.068 2010 4.7 4.6 4.8 0.068 0.067 0.071 2015 5.5 5.3 5.8 0.077 0.075 0.082 2020 6.9 6.6 7.4 0.095 0.091 0.104 2025 8.4 8.0 9.3 0.117 0.110 0.131 2030 9.8 9.2 0.138 0.127 2035 11.4 10.4 0.160 0.145 2040 12.9 11.6 0.182 0.162 2045 15.1 13.5 0.217 0.190 2050 18.2 16.2 0.270 0.235 Table A55
Total Dependency Ratio Egypt
Fertility: High Med Low High Med Low High Med Low Mortality: High High High Med Med Med Low Low Low
1995 0.730 0.730 0.730 0.730 0.730 0.730 0.730 0.730 0.730 2000 0.662 0.651 0.641 0.662 0.651 0.641 0.662 0.651 0.641 2005 0.597 0.570 0.543 0.598 0.571 0.544 0.600 0.573 0.546 2010 0.561 0.512 0.462 0.565 0.515 0.465 0.570 0.520 0.469 2015 0.531 0.468 0.403 0.537 0.473 0.407 0.549 0.481 0.415 2020 0.523 0.453 0.379 0.529 0.459 0.384 0.551 0.469 0.395 2025 0.527 0.456 0.381 0.535 0.464 0.389 0.570 0.478 0.404 2030 0.531 0.461 0.390 0.541 0.472 0.402 2035 0.531 0.461 0.393 0.545 0.476 0.408 2040 0.528 0.458 0.392 0.545 0.477 0.413 2045 0.536 0.469 0.409 0.557 0.493 0.436 2050 0.557 0.499 0.452 0.583 0.529 0.488 Table B1. Summary of labor force and labor force participation rates Asia
Labor force (1000s) Labor force participation rates (%) Year 15-64 65 + Combined 15-64 65 + Combined Male 1950 400,124 16,201 416,325 93.3 63.8 91.6 1960 453,736 19,584 473,320 92.4 58.9 90.2 1970 553,124 21,647 574,771 90.5 54.3 88.3 1980 698,362 24,853 723,215 88.9 46.8 86.2 1990 886,650 31,277 917,927 87.7 43.5 84.8 2000 1,054,949 37,915 1,092,864 87.0 38.4 83.3 2010 1,227,049 43,775 1,270,824 85.8 34.2 81.6 2020 1,364,996 63,524 1,428,520 86.2 34.7 80.9 2030 1,443,559 88,504 1,532,063 86.0 34.1 79.0 2040 1,473,355 117,572 1,590,927 86.1 33.3 77.1 2050 1,475,140 132,203 1,607,344 85.8 31.9 75.3 Female 1950 233,855 5,569 239,424 58.1 17.4 55.1 1960 271,014 6,009 277,023 57.9 15.9 54.7 1970 340,800 6,713 347,513 58.5 14.4 55.2 1980 446,173 7,921 454,094 59.9 12.5 56.1 1990 582,832 11,428 594,260 60.8 13.4 56.9 2000 713,900 14,933 728,832 62.0 12.7 57.4 2010 848,267 18,695 866,961 62.4 12.2 57.3 2020 932,894 26,848 959,742 61.9 12.3 55.6 2030 978,050 37,461 1,015,511 61.1 12.1 53.2 2040 993,960 49,562 1,043,521 61.0 11.7 50.9 2050 991,581 55,348 1,046,929 60.5 11.1 48.9 Total 1950 633,979 21,770 655,749 76.2 37.9 73.7 1960 724,750 25,593 750,343 75.5 36.0 72.8 1970 893,924 28,360 922,284 74.9 32.7 72.0 1980 1,144,535 32,774 1,177,309 74.8 28.1 71.5 1990 1,469,482 42,700 1,512,182 74.7 27.3 71.2 2000 1,768,848 52,848 1,821,696 74.8 24.4 70.6 2010 2,075,316 62,469 2,137,785 74.4 22.2 69.6 2020 2,297,890 90,372 2,388,262 74.4 22.5 68.4 2030 2,421,609 125,965 2,547,574 73.9 22.1 66.2 2040 2,467,315 167,133 2,634,448 73.9 21.5 64.0 2050 2,466,721 187,551 2,654,272 73.4 20.5 62.1 Table B2. Summary of labor force and labor force participation rates East Asia
Labor force (1000s) Labor force participation rates (%) Year 15-64 65 + Combined 15-64 65 + Combined Male 1950 200,089 7,065 207,154 94.1 56.2 92.0 1960 216,672 8,893 225,565 93.4 51.2 90.5 1970 266,894 9,027 275,921 91.8 46.4 89.0 1980 333,414 9,432 342,846 90.3 36.2 86.7 1990 415,431 12,479 427,910 88.8 33.9 84.8 2000 463,201 14,701 477,902 88.7 28.6 83.3 2010 502,191 15,692 517,883 86.8 23.9 80.4 2020 512,556 23,040 535,597 86.7 24.3 78.1 2030 498,019 29,991 528,009 85.2 23.6 74.2 2040 470,496 38,778 509,274 85.9 23.0 71.1 2050 448,992 36,699 485,691 85.2 21.2 69.4 Female 1950 130,825 2,051 132,876 65.6 11.8 61.3 1960 146,985 2,311 149,296 66.8 10.9 61.9 1970 187,539 2,568 190,107 68.5 10.1 63.5 1980 246,897 2,730 249,627 71.7 7.9 65.9 1990 334,962 4,551 339,513 76.4 9.9 70.1 2000 381,529 5,861 387,390 77.3 9.3 69.6 2010 415,919 7,029 422,948 76.0 8.8 67.4 2020 409,500 10,324 419,824 73.4 9.0 62.4 2030 374,270 13,528 387,798 68.1 8.7 55.0 2040 332,332 17,446 349,778 65.2 8.4 48.8 2050 291,083 16,351 307,434 59.5 7.6 43.6 Total 1950 330,914 9,116 340,030 80.3 30.4 76.9 1960 363,657 11,204 374,861 80.5 29.1 76.4 1970 454,433 11,595 466,028 80.5 25.8 76.5 1980 580,311 12,162 592,473 81.3 20.1 76.5 1990 750,393 17,043 767,436 82.8 20.6 77.6 2000 844,730 20,562 865,292 83.2 18.0 76.6 2010 918,110 22,720 940,831 81.6 15.6 74.0 2020 922,057 33,364 955,421 80.3 15.9 70.3 2030 872,289 43,519 915,808 76.9 15.4 64.6 2040 802,828 56,224 859,052 75.9 14.9 59.9 2050 740,074 53,051 793,125 72.8 13.6 56.4 Table B3. Summary of labor force and labor force participation rates Southeast Asia
Labor force (1000s) Labor force participation rates (%) Year 15-64 65 + Combined 15-64 65 + Combined Male 1950 48,198 2,289 50,487 92.6 74.2 91.6 1960 56,349 2,346 58,695 91.3 69.6 90.2 1970 67,169 2,655 69,824 89.0 63.1 87.7 1980 85,668 3,330 88,998 87.1 58.3 85.5 1990 111,714 4,185 115,899 85.8 53.3 84.0 2000 140,741 5,166 145,907 85.5 47.8 83.1 2010 169,097 6,115 175,212 85.2 42.4 82.3 2020 194,260 8,680 202,940 85.9 42.7 82.4 2030 208,711 13,587 222,299 86.0 42.9 81.0 2040 215,320 19,079 234,398 85.8 41.7 79.0 2050 216,985 23,411 240,397 85.7 40.2 77.2 Female 1950 27,012 1,166 28,178 52.2 31.6 50.9 1960 32,490 1,238 33,728 52.0 30.0 50.7 1970 41,806 1,395 43,201 53.9 27.2 52.3 1980 58,610 1,856 60,466 57.7 25.9 55.6 1990 81,728 2,537 84,265 61.6 25.9 59.2 2000 106,336 3,303 109,639 64.0 24.1 61.0 2010 131,933 4,177 136,110 66.5 22.5 62.8 2020 149,046 5,884 154,930 66.4 22.6 61.9 2030 158,012 9,094 167,106 66.1 22.6 59.8 2040 161,699 12,393 174,092 65.7 21.6 57.4 2050 161,831 15,021 176,852 65.5 20.6 55.3 Total 1950 75,210 3,455 78,665 72.5 51.0 71.2 1960 88,839 3,584 92,423 71.6 47.8 70.2 1970 108,975 4,050 113,025 71.3 43.4 69.6 1980 144,278 5,186 149,464 72.2 40.3 70.2 1990 193,442 6,681 200,123 73.6 37.8 71.4 2000 247,077 8,469 255,546 74.7 34.6 71.9 2010 301,030 10,291 311,321 75.9 31.2 72.4 2020 343,306 14,564 357,870 76.2 31.4 72.0 2030 366,723 22,681 389,404 76.1 31.5 70.3 2040 377,019 31,472 408,490 75.8 30.6 68.1 2050 378,816 38,433 417,249 75.7 29.3 66.1 Table B4. Summary of labor force and labor force participation rates South Asia
Labor force (1000s) Labor force participation rates (%) Year 15-64 65 + Combined 15-64 65 + Combined Male 1950 138,848 6,238 145,086 92.4 71.3 91.3 1960 164,349 7,633 171,982 91.6 67.2 90.1 1970 199,015 9,062 208,077 89.7 62.3 88.0 1980 252,680 11,225 263,905 88.2 58.6 86.3 1990 322,352 13,454 335,806 87.5 55.3 85.5 2000 402,834 16,566 419,400 86.3 51.0 84.0 2010 495,687 19,959 515,645 85.6 47.1 83.0 2020 585,352 28,373 613,725 86.5 47.5 83.4 2030 651,556 41,517 693,073 87.2 47.2 83.0 2040 693,632 56,171 749,804 87.2 46.1 81.8 2050 708,320 73,076 781,397 87.0 45.4 80.1 Female 1950 68,818 1,965 70,783 50.2 20.3 48.2 1960 83,401 2,050 85,451 49.6 18.5 47.7 1970 102,013 2,375 104,388 48.9 16.8 46.8 1980 128,772 2,991 131,763 47.9 15.5 45.8 1990 150,776 3,865 154,641 43.7 14.9 41.7 2000 217,787 5,073 222,860 49.8 14.1 47.1 2010 287,208 6,617 293,825 52.8 13.7 49.6 2020 336,058 8,623 344,681 52.6 12.7 48.7 2030 370,166 12,605 382,771 52.1 12.7 47.3 2040 389,904 16,996 406,899 51.3 12.3 45.3 2050 395,386 22,097 417,482 50.5 12.1 43.2 Total 1950 207,666 8,203 215,869 72.3 44.5 70.6 1960 247,750 9,683 257,433 71.3 43.2 69.6 1970 301,028 11,437 312,465 69.9 39.9 68.0 1980 381,452 14,216 395,668 68.7 37.0 66.7 1990 473,128 17,496 490,624 66.3 34.8 64.3 2000 620,621 21,640 642,260 68.6 31.6 66.0 2010 782,895 26,576 809,471 69.7 29.3 66.7 2020 921,410 36,995 958,406 70.0 29.0 66.4 2030 1,021,722 54,122 1,075,845 70.1 28.9 65.4 2040 1,083,536 73,167 1,156,703 69.7 28.2 63.7 2050 1,103,706 95,173 1,198,879 69.1 27.6 61.8 Table B5. Summary of labor force and labor force participation rates Japan
Labor force (1000s) Labor force participation rates (%) Year 15-64 65 + Combined 15-64 65 + Combined Male 1950 21,205 947 22,152 87.6 54.5 85.4 1960 25,763 1,277 27,040 87.7 54.5 85.3 1970 30,594 1,764 32,358 87.0 54.5 84.3 1980 33,497 2,047 35,544 86.1 45.8 82.0 1990 36,249 2,362 38,611 84.2 39.6 78.8 2000 36,900 3,029 39,930 85.0 33.4 76.1 2010 34,575 3,194 37,770 84.5 27.2 71.8 2020 31,562 3,669 35,230 84.1 26.1 68.3 2030 29,728 3,362 33,090 83.7 24.2 67.0 2040 26,385 3,671 30,056 83.3 25.0 64.8 2050 24,184 3,497 27,681 83.2 23.9 63.3 Female 1950 13,370 518 13,888 52.2 21.6 49.5 1960 16,696 641 17,337 54.0 21.0 51.0 1970 19,936 815 20,751 54.3 19.7 50.8 1980 20,743 964 21,707 52.1 15.8 47.3 1990 24,146 1,376 25,522 56.2 15.6 49.3 2000 26,426 1,808 28,234 61.6 14.4 50.9 2010 26,510 2,100 28,610 65.9 13.4 51.2 2020 24,330 2,360 26,690 66.5 12.8 48.5 2030 22,464 2,149 24,613 65.1 11.7 46.6 2040 19,830 2,266 22,096 64.8 11.9 44.5 2050 18,361 2,137 20,498 65.6 11.4 43.9 Total 1950 34,575 1,465 36,040 69.4 35.4 66.8 1960 42,459 1,918 44,377 70.4 35.5 67.6 1970 50,530 2,579 53,109 70.3 35.0 67.0 1980 54,240 3,011 57,251 68.9 28.5 64.1 1990 60,395 3,738 64,133 70.2 25.2 63.6 2000 63,327 4,837 68,164 73.3 22.4 63.1 2010 61,086 5,294 66,379 75.3 19.3 61.2 2020 55,892 6,029 61,921 75.4 18.6 58.1 2030 52,191 5,512 57,703 74.5 17.1 56.4 2040 46,216 5,937 52,152 74.2 17.6 54.3 2050 42,545 5,633 48,178 74.5 16.9 53.3 Table B6. Summary of labor force and labor force participation rates South Korea
Labor force (1000s) Labor force participation rates (%) Year 15-64 65 + Combined 15-64 65 + Combined Male 1950 5,163 172 5,335 90.5 66.7 89.4 1960 5,756 179 5,935 85.8 60.5 84.8 1970 7,237 161 7,398 83.2 40.0 81.3 1980 9,285 229 9,514 77.6 38.2 75.7 1990 11,615 293 11,908 77.4 36.3 75.3 2000 13,617 386 14,003 79.8 32.4 76.7 2010 14,486 543 15,029 80.1 29.3 75.3 2020 14,717 748 15,465 79.0 28.6 72.8 2030 14,079 1,144 15,223 78.9 28.6 69.7 2040 13,146 1,376 14,522 78.8 27.0 66.7 2050 12,495 1,360 13,855 78.7 25.4 65.3 Female 1950 1,519 29 1,548 27.4 8.0 26.2 1960 2,004 49 2,053 28.7 9.2 27.3 1970 3,522 68 3,590 40.3 10.5 38.2 1980 5,892 118 6,010 50.2 13.8 47.7 1990 7,485 240 7,725 51.1 17.9 48.3 2000 9,563 314 9,878 57.8 16.0 53.4 2010 11,133 403 11,535 64.1 14.5 57.3 2020 11,072 527 11,599 62.6 14.0 54.1 2030 10,311 786 11,097 61.9 14.1 49.9 2040 9,513 915 10,429 61.8 13.2 46.7 2050 9,055 879 9,934 62.3 12.0 45.5 Total 1950 6,682 201 6,883 59.4 32.4 57.9 1960 7,760 228 7,988 56.7 27.4 55.0 1970 10,759 229 10,988 61.7 21.8 59.4 1980 15,177 347 15,524 64.0 23.9 61.7 1990 19,100 533 19,633 64.4 24.9 61.8 2000 23,181 700 23,881 68.9 22.2 64.9 2010 25,619 946 26,564 72.3 20.4 66.3 2020 25,789 1,275 27,064 71.0 20.0 63.4 2030 24,390 1,931 26,320 70.7 20.2 59.7 2040 22,659 2,292 24,951 70.6 19.1 56.6 2050 21,550 2,239 23,789 70.9 17.7 55.2 Table B7. Summary of labor force and labor force participation rates The Philippines
Labor force (1000s) Labor force participation rates (%) Year 15-64 65 + Combined 15-64 65 + Combined Male 1950 5,005 199 5,204 91.7 70.7 90.7 1960 6,440 217 6,657 89.2 67.7 88.3 1970 8,435 311 8,746 86.4 64.8 85.4 1980 11,280 391 11,671 84.5 61.9 83.5 1990 14,487 572 15,059 84.6 59.0 83.2 2000 18,941 687 19,628 83.2 54.5 81.7 2010 24,211 902 25,113 83.2 50.1 81.3 2020 29,726 1,429 31,155 84.5 50.5 82.0 2030 34,042 2,168 36,210 86.5 49.7 82.8 2040 36,612 2,921 39,533 86.9 48.3 82.0 2050 38,002 3,936 41,938 86.6 47.7 80.4 Female 1950 2,296 80 2,376 40.8 17.0 39.0 1960 3,075 102 3,177 42.7 19.7 41.1 1970 4,261 120 4,381 43.9 22.6 42.8 1980 6,161 180 6,341 46.0 25.5 45.0 1990 8,347 325 8,672 49.2 28.4 47.9 2000 11,567 393 11,960 51.3 26.2 49.7 2010 15,561 512 16,073 54.5 24.1 52.4 2020 18,916 795 19,711 54.9 24.2 52.3 2030 21,389 1,224 22,613 55.7 23.7 51.9 2040 22,702 1,699 24,401 55.7 22.7 50.6 2050 23,357 2,261 25,617 55.4 21.9 48.8 Total 1950 7,301 279 7,580 65.9 37.0 64.0 1960 9,515 319 9,834 65.9 38.0 64.4 1970 12,696 431 13,127 65.2 42.7 64.1 1980 17,441 571 18,012 65.2 42.7 64.2 1990 22,834 845 23,679 67.0 40.0 65.4 2000 30,508 1,080 31,588 67.3 39.2 65.7 2010 39,772 1,414 41,186 69.0 36.0 66.9 2020 48,642 2,223 50,865 69.9 36.4 67.2 2030 55,432 3,391 58,823 71.3 35.6 67.4 2040 59,313 4,620 63,933 71.5 34.1 66.3 2050 61,358 6,196 67,555 71.3 33.4 64.6 Table B8. Summary of labor force and labor force participation rates Thailand
Labor force (1000s) Labor force participation rates (%) Year 15-64 65 + Combined 15-64 65 + Combined Male 1950 5,048 154 5,202 92.0 56.0 90.3 1960 6,366 170 6,536 91.5 52.0 89.8 1970 8,091 227 8,318 90.2 48.0 88.1 1980 11,762 320 12,082 89.3 44.1 87.0 1990 15,862 425 16,287 89.4 40.1 86.6 2000 19,043 581 19,624 89.8 37.5 86.3 2010 21,227 745 21,972 90.0 35.1 85.5 2020 22,300 1,095 23,395 89.7 35.4 83.7 2030 22,128 1,686 23,814 88.8 35.4 80.3 2040 20,991 2,262 23,253 88.6 34.2 76.7 2050 19,859 2,417 22,275 88.5 32.2 74.3 Female 1950 4,555 115 4,670 84.2 34.7 81.3 1960 5,717 123 5,840 66.0 30.9 64.4 1970 7,448 164 7,612 81.1 27.2 77.7 1980 10,527 214 10,741 79.7 23.2 76.0 1990 13,899 261 14,160 78.6 19.3 74.4 2000 16,498 351 16,849 78.0 17.3 72.7 2010 18,240 446 18,685 77.2 15.7 70.6 2020 18,878 646 19,523 75.9 15.8 67.4 2030 18,481 972 19,454 74.4 15.8 62.8 2040 17,453 1,277 18,730 74.1 15.2 58.6 2050 16,500 1,355 17,855 74.2 14.2 56.2 Total 1950 9,603 269 9,872 88.1 44.4 85.8 1960 12,083 293 12,376 77.4 40.4 75.7 1970 15,539 391 15,930 85.6 36.3 82.8 1980 22,289 534 22,823 84.5 32.4 81.4 1990 29,761 686 30,447 84.0 28.5 80.4 2000 35,541 932 36,473 84.0 26.1 79.4 2010 39,467 1,191 40,657 83.6 24.0 77.9 2020 41,178 1,741 42,918 82.8 24.3 75.4 2030 40,609 2,658 43,268 81.6 24.4 71.3 2040 38,444 3,539 41,983 81.4 23.5 67.4 2050 36,359 3,772 40,131 81.4 22.1 65.0 Table B9. Summary of labor force and labor force participation rates Indonesia
Labor force (1000s) Labor force participation rates (%) Year 15-64 65 + Combined 15-64 65 + Combined Male 1950 21,148 1,327 22,475 92.9 88.0 92.6 1960 24,753 1,259 26,012 91.7 84.0 91.3 1970 28,861 1,247 30,108 88.9 74.1 88.2 1980 35,548 1,501 37,049 85.8 65.0 84.7 1990 46,074 1,855 47,929 84.1 56.6 82.6 2000 57,728 2,202 59,930 84.5 48.5 82.3 2010 68,679 2,508 71,187 85.2 40.4 82.0 2020 78,236 3,328 81,563 86.0 40.2 82.2 2030 83,414 5,006 88,421 86.0 40.4 80.9 2040 85,359 7,132 92,490 85.7 39.8 78.7 2050 85,844 8,662 94,506 85.7 38.3 77.0 Female 1950 6,866 506 7,372 30.6 30.8 30.6 1960 8,747 489 9,236 32.0 28.6 31.8 1970 12,341 482 12,823 37.1 24.1 36.3 1980 19,393 687 20,080 45.6 25.1 44.3 1990 28,992 953 29,945 52.0 25.2 50.3 2000 39,927 1,316 41,243 57.9 24.1 55.5 2010 51,470 1,794 53,264 63.8 23.1 60.2 2020 57,999 2,388 60,387 64.0 22.8 59.7 2030 61,266 3,539 64,806 63.8 22.7 58.0 2040 62,303 4,953 67,255 63.4 22.0 55.7 2050 62,177 6,011 68,189 63.3 20.8 53.7 Total 1950 28,014 1,833 29,847 61.9 58.2 61.7 1960 33,500 1,748 35,248 61.6 54.5 61.2 1970 41,202 1,729 42,931 62.7 46.9 61.8 1980 54,941 2,188 57,129 65.4 43.4 64.2 1990 75,066 2,808 77,874 67.9 39.8 66.2 2000 97,655 3,518 101,173 71.2 35.2 68.7 2010 120,149 4,302 124,452 74.5 30.8 71.0 2020 136,235 5,715 141,951 75.0 30.5 70.8 2030 144,681 8,546 153,226 75.0 30.6 69.3 2040 147,662 12,084 159,746 74.6 29.9 67.0 2050 148,022 14,673 162,695 74.6 28.5 65.1 Table B10. Summary of labor force and labor force participation rates India
Labor force (1000s) Labor force participation rates (%) Year 15-64 65 + Combined 15-64 65 + Combined Male 1950 99,290 3,817 103,107 92.2 71.6 91.3 1960 118,549 5,139 123,688 91.4 67.5 90.1 1970 143,991 6,581 150,572 89.9 63.6 88.3 1980 181,362 8,468 189,830 88.6 59.9 86.7 1990 230,143 10,158 240,301 88.0 56.5 86.0 2000 282,125 12,614 294,739 86.9 52.7 84.5 2010 339,267 15,221 354,488 85.9 48.9 83.2 2020 392,255 21,275 413,530 86.9 49.1 83.6 2030 423,865 30,672 454,537 87.6 48.7 83.1 2040 440,865 40,414 481,279 87.3 47.4 81.6 2050 443,438 50,407 493,845 87.0 46.5 79.9 Female 1950 51,114 1,461 52,575 51.7 22.0 49.9 1960 61,478 1,497 62,975 50.6 19.8 48.8 1970 73,919 1,745 75,664 49.3 17.6 47.3 1980 91,109 2,114 93,223 47.8 15.4 45.6 1990 103,038 2,706 105,744 42.5 14.3 40.4 2000 134,264 3,585 137,849 44.6 13.5 42.1 2010 170,896 4,622 175,518 46.6 12.9 43.6 2020 197,242 6,279 203,522 46.6 12.7 43.1 2030 212,578 8,953 221,531 46.5 12.6 42.0 2040 220,818 11,805 232,623 45.9 12.3 40.3 2050 222,124 14,800 236,924 45.4 12.0 38.7 Total 1950 150,404 5,278 155,682 72.9 44.1 71.3 1960 180,027 6,636 186,663 71.7 43.7 70.1 1970 217,910 8,326 226,236 70.2 41.1 68.4 1980 272,471 10,582 283,053 68.9 37.9 66.9 1990 333,181 12,815 345,996 66.1 34.7 64.0 2000 416,389 16,199 432,588 66.6 32.1 64.0 2010 510,163 19,843 530,005 67.0 29.6 64.0 2020 589,497 27,555 617,052 67.4 29.7 63.8 2030 636,443 39,625 676,068 67.7 29.6 62.9 2040 661,683 52,219 713,902 67.1 28.8 61.2 2050 665,563 65,206 730,769 66.7 28.2 59.4 Table B11. Summary of labor force and labor force participation rates Bangladesh
Labor force (1000s) Labor force participation rates (%) Year 15-64 65 + Combined 15-64 65 + Combined Male 1950 12,908 684 13,592 94.9 90.0 94.6 1960 14,415 908 15,323 94.3 87.4 93.8 1970 16,590 1,064 17,654 92.7 82.3 92.0 1980 20,834 1,123 21,957 90.9 73.0 89.7 1990 26,011 1,107 27,118 88.2 65.2 87.0 2000 35,802 1,217 37,019 87.7 59.0 86.3 2010 45,065 1,523 46,588 89.2 52.7 87.3 2020 53,782 2,230 56,012 90.0 52.6 87.5 2030 61,279 3,088 64,367 90.8 51.6 87.7 2040 64,861 4,742 69,603 91.3 52.2 86.9 2050 64,895 7,390 72,285 91.0 51.9 84.5 Female 1950 7,993 263 8,256 73.1 34.9 70.6 1960 9,459 301 9,760 72.0 34.6 69.6 1970 11,507 360 11,867 71.2 34.4 68.9 1980 15,193 513 15,706 70.2 34.2 67.8 1990 18,497 579 19,076 66.0 33.3 64.1 2000 26,274 686 26,960 67.7 32.9 65.9 2010 33,692 915 34,607 69.8 32.5 67.7 2020 39,955 1,417 41,372 70.0 32.9 67.4 2030 44,820 2,121 46,941 69.8 32.3 66.4 2040 47,099 3,234 50,333 69.6 32.2 64.8 2050 47,267 4,921 52,188 69.3 31.7 62.3 Total 1950 20,901 947 21,848 85.1 62.5 83.8 1960 23,874 1,209 25,083 84.0 63.3 82.7 1970 28,097 1,424 29,521 82.5 60.9 81.1 1980 36,027 1,636 37,663 80.8 53.9 79.1 1990 44,508 1,687 46,195 77.4 49.1 75.8 2000 62,076 1,903 63,979 77.9 45.9 76.3 2010 78,756 2,438 81,194 79.7 42.7 77.7 2020 93,737 3,647 97,384 80.2 42.6 77.7 2030 106,099 5,209 111,308 80.6 41.5 77.2 2040 111,960 7,975 119,936 80.7 41.7 76.0 2050 112,162 12,310 124,473 80.4 41.3 73.5 Table B12. Summary of labor force and labor force participation rates Egypt
Labor force (1000s) Labor force participation rates (%) Year 15-64 65 + Combined 15-64 65 + Combined Male 1950 5,753 217 5,970 92.8 75.3 92.0 1960 6,826 256 7,082 90.3 62.0 88.8 1970 8,295 335 8,630 86.1 48.7 83.6 1980 10,449 280 10,729 83.5 35.4 80.7 1990 13,448 290 13,738 83.4 28.6 80.4 2000 17,328 304 17,632 82.1 24.6 78.9 2010 22,341 341 22,682 83.1 21.5 79.6 2020 26,590 576 27,165 84.3 22.3 79.6 2030 29,604 858 30,462 85.5 21.6 78.9 2040 31,633 1,111 32,744 85.0 20.9 77.0 2050 32,165 1,574 33,739 84.5 21.0 74.1 Female 1950 1,543 54 1,597 24.4 15.0 23.9 1960 1,957 70 2,027 25.9 13.9 25.2 1970 2,614 101 2,715 27.5 12.1 26.2 1980 3,571 94 3,665 29.3 9.8 27.9 1990 4,976 109 5,085 31.9 8.9 30.3 2000 7,562 125 7,688 37.2 8.0 35.1 2010 11,206 143 11,349 43.2 7.0 40.6 2020 13,110 224 13,335 43.0 7.0 39.6 2030 14,279 339 14,618 42.5 7.0 38.1 2040 14,812 458 15,270 40.9 7.0 35.8 2050 14,944 642 15,585 40.3 7.0 33.8 Total 1950 7,296 271 7,567 58.3 41.9 57.5 1960 8,783 326 9,109 58.1 35.6 56.9 1970 10,909 436 11,345 56.9 28.6 54.9 1980 14,020 374 14,394 56.7 21.3 54.4 1990 18,424 399 18,823 58.1 18.5 55.6 2000 24,891 429 25,320 60.0 15.3 57.2 2010 33,547 484 34,031 63.5 13.4 60.3 2020 39,700 800 40,500 64.0 13.9 59.7 2030 43,883 1,198 45,081 64.4 13.6 58.6 2040 46,445 1,569 48,014 63.3 13.3 56.4 2050 47,108 2,216 49,324 62.7 13.3 53.8 Table 1. Summary measures of aging for Asia, major regions, seven ANE countries. Growth rate of Elderly population (1000s) Elderly Population Old age dependency ratio 2000 2025 2050 2000-25 2025-50 2000 2025 2050
Asia 207,349 454,964 864,614 3.14 2.57 0.086 0.144 0.250 East Asia 114,390 241,217 389,089 2.98 1.91 0.113 0.210 0.383 Southeast Asia 24,503 58,253 131,138 3.46 3.25 0.074 0.124 0.262 South Asia 68,457 155,494 344,388 3.28 3.18 0.076 0.112 0.216
South Korea 3,152 8,020 12,665 3.74 1.83 0.094 0.226 0.417 Indonesia 10,001 23,078 51,500 3.34 3.21 0.073 0.123 0.260 Phiippines 2,758 7,786 18,558 4.15 3.47 0.061 0.105 0.216 Thailand 3,576 8,924 17,077 3.66 2.60 0.084 0.178 0.382 Egypt 2,814 7,331 16,604 3.83 3.27 0.052 0.084 0.213 Bangladesh 4,149 10,494 29,787 3.71 4.17 0.081 0.123 0.232 India 50,466 111,934 231,266 3.19 2.90 0.068 0.112 0.221
Source: UN Population Division, 1998. Note: All data employ the medium fertility variant. Table 2. Fertility assumptions, medium UN projections.
Country 1990-95 1995-00 2000-05 2005-10 2010-15 2015-2020 South Korea 1.65 1.69 1.76 1.83 1.89 1.90 Indonesia 2.58 2.26 2.10 2.10 2.10 2.10 Philippines 3.62 3.19 2.76 2.33 2.10 2.10 Thailand 1.74 1.74 1.78 1.85 1.90 1.90 Egypt 3.40 2.88 2.36 2.10 2.10 2.10 Bangladesh 3.11 2.82 2.53 2.24 2.10 2.10 India 3.13 2.72 2.31 2.10 2.10 2.10
Source: UN 1998. Note: For countries included in this table TFR is constant after 2015-2020. Table 3. Projected life expectancy, study countries. Projected Country 1990-95 2000-05 2010-15 2020-25 2030-35 2040-45 Increase Republic of Korea 70.9 73.5 75.4 77.1 78.4 79.4 8.5 Bangladesh 55.6 60.7 65.3 69.2 71.9 74.1 18.5 Indonesia 62.6 67.3 70.5 72.9 74.9 76.6 14.0 Thailand 68.8 69.4 72.8 75.4 76.0 76.6 7.8 Egypt 63.9 68.3 71.3 73.7 75.6 77.2 13.3 India 60.3 64.2 67.3 70.4 72.7 74.2 13.9 Philippines 66.3 69.8 72.3 74.4 76.3 77.7 11.4
Source: UN Population Division, 1998. Table 4. Total labor force, labor force aged 65 and over, regions of Asia and eight ANE countries
1950 2000 2050 1950 2000 2050
Asia 21,769 52,848 187,551 709,562 1,860,155 2,672,600 East Asia 9,115 20,562 53,051 366,935 874,564 853,782 Southeast Asia 3,456 8,469 45,054 84,805 260,621 426,501 South Asia 8,203 21,640 95,173 235,272 650,951 1,197,792
Japan 1,465 4,837 5,633 36,650 68,164 48,178 South Korea 201 700 2,239 7,449 23,881 23,789 The Philippines 279 1,080 6,196 8,194 32,067 67,555 Thailand 269 932 3,772 10,894 37,173 40,340 Indonesia 1,834 3,518 14,673 32,149 102,856 163,584 India 5,278 16,199 65,206 169,907 445,770 738,140 Bangladesh 946 1,903 12,310 24,153 68,471 127,918 Egypt 271 429 2,216 8,246 26,081 49,668
Source: ILO, 1996. Calculations by authors. Table 5. Demographic characteristics of the older population and labor force, Asia
55-59 60-64 65-74 75+ Total Age distribution of men 55 and older (percent) 2000 29.0 24.7 33.2 13.1 100.0 2025 29.8 24.0 31.8 14.4 100.0 2050 22.5 21.9 32.5 23.1 100.0
2000 26.2 23.0 33.4 17.3 100.0 2025 27.1 22.4 31.8 18.6 100.0 2050 19.7 19.6 31.3 29.4 100.0 Sex ratio (100 x men/women) 2000 101.9 98.8 91.6 69.3 92.1 2025 100.9 98.0 91.8 71.1 91.8 2050 103.9 101.2 94.0 71.5 90.7 Age distribution of the labor force (percent) 2000 43.7 27.8 28.5 100.0 2025 44.2 27.5 28.5 100.0 2050 37.4 27.1 35.5 100.0 Sex ratio of the labor force 2000 191.9 209.3 255.9 212.8 2025 187.8 199.1 213.6 197.8 2050 198.9 207.8 221.6 209.1
Sources: UN, 1999; ILO, 1996; calculations by authors. Table 6. Demographic characteristics of the older population, Asia and ANE countries
Percentage 55 and older Sex ratio, population 55+ Percentage 55+ in labor force 2000 2025 2050 2000 2025 2050 2000 2025 2050 Asia 12.1 20.1 29.8 92.1 91.8 90.7 38.3 31.9 31.8 East Asia 16.8 18.4 33.6 92.2 90.5 88.8 36.9 33.8 28.4 Southeast Asia 13.9 14.1 24.2 84.8 86.6 87.4 51.4 52.8 47.3 South Asia 15.4 16.6 26.6 92.1 91.5 90.7 45.8 44.3 41.1
Japan 22.9 37.5 43.1 81.6 82.7 83.7 40.7 33.8 29.2 South Korea 13.9 17.6 34.7 77.5 82.7 83.5 43.0 39.1 31.4 Philippines 13.4 13.9 21.0 90.9 91.6 89.6 55.9 53.4 49.6 Thailand 16.2 17.1 31.2 83.6 85.1 84.6 47.5 45.2 37.4 Indonesia 12.1 13.7 23.4 87.4 87.4 86.0 52.7 50.8 44.5 India 13.7 15.6 21.6 94.1 95.6 93.9 45.7 43.7 40.8 Bangladesh 12.2 12.4 17.0 106.8 96.6 96.3 65.0 64.9 57.3 Egypt 13.4 13.0 19.4 84.9 90.9 89.8 33.2 32.3 29.6 Sources: UN, 1999; ILO, 1996; calculations by authors. Table 7. Summary of Dependency Ratios and Economic Support Ratios, Medium Variant,Asia and ANE Countries, 2000, 2025, and 2050.
Total dependency ratio Child dependency ratio Old age dependency ratio Economic support ratio 2000 2025 2050 2000 2025 2050 2000 2025 2050 2000 2025 2050
East Asia 0.462 0.474 0.649 0.349 0.265 0.266 0.113 0.210 0.383 0.359 0.353 0.428 Southeast Asia 0.568 0.460 0.570 0.494 0.336 0.308 0.074 0.124 0.262 0.430 0.356 0.403 South Asia 0.649 0.472 0.522 0.573 0.360 0.306 0.076 0.112 0.216 0.476 0.365 0.381
Japan 0.468 0.673 0.838 0.217 0.226 0.254 0.250 0.447 0.583 0.344 0.431 0.490 South Korea 0.393 0.477 0.678 0.299 0.252 0.270 0.094 0.226 0.417 0.316 0.353 0.437 Philippines 0.676 0.458 0.521 0.615 0.353 0.305 0.061 0.105 0.216 0.494 0.357 0.381 Thailand 0.450 0.453 0.660 0.366 0.274 0.278 0.084 0.178 0.382 0.355 0.344 0.434 Indonesia 0.546 0.456 0.573 0.473 0.333 0.313 0.073 0.123 0.260 0.417 0.354 0.405 Bangladesh 0.622 0.428 0.523 0.569 0.344 0.309 0.052 0.084 0.213 0.465 0.341 0.382 India 0.620 0.459 0.531 0.540 0.336 0.300 0.081 0.123 0.232 0.459 0.356 0.384 Egypt 0.651 0.464 0.529 0.584 0.352 0.308 0.068 0.112 0.221 0.479 0.360 0.385
Source: See text. Table 8. Projected life expectancy among the study countries, Asia and the Near East, 1995-2050
Average value Rate of increase (years per quinquennia) (1995-2050) Projected Model Difference
South Korea 76.7 0.74 0.93 -0.19 Bangladesh 68.1 1.69 1.54 0.15 Indonesia 72.3 1.23 1.11 0.13 Thailand 73.9 0.81 0.95 -0.14 Egypt 73.1 1.16 1.05 0.11 India 69.6 1.23 1.38 -0.15 Philippines 74.0 1.01 0.96 0.04 Table 9. Popuation 65 and older, three mortality variants, regions of Asia and seven ANE countries
Population Population 65+ in 2025 Population 65+ in 2050 65+ in 2000 Low Medium High Medium High
Asia 207,301 496,731 454,718 428,612 863,601 766,636 East Asia 114,390 261,742 241,217 228,807 389,089 346,968 Southeast Asia 24,503 64,773 58,253 54,599 131,138 113,214 South Asia 68,457 172,367 155,494 144,986 344,388 299,540
South Korea 3,152 8,820 8,020 7,562 12,665 10,967 Indonesia 10,001 26,042 23,078 21,728 51,500 43,595 Phiippines 2,758 8,735 7,786 7,297 18,558 15,796 Thailand 3,576 9,541 8,924 8,346 17,077 15,542 Egypt 2,814 8,305 7,331 6,829 16,604 14,041 Bangladesh 4,149 11,874 10,494 9,776 29,787 25,120 India 50,466 123,738 111,934 104,412 231,266 201,797 Table 10. Growth rate, popuation 65 and older, three mortality variants, regions of Asia and seven ANE countries
Growth rate 2000-2025 Growth rate 2025-50 Low Medium High Medium High
Asia 3.50 3.14 2.91 2.57 2.33 East Asia 3.31 2.98 2.77 1.91 1.67 Southeast Asia 3.89 3.46 3.20 3.25 2.92 South Asia 3.69 3.28 3.00 3.18 2.90
South Korea 4.12 3.74 3.50 1.83 1.49 Indonesia 3.83 3.34 3.10 3.21 2.79 Philippines 4.61 4.15 3.89 3.47 3.09 Thailand 3.93 3.66 3.39 2.60 2.49 Egypt 4.33 3.83 3.55 3.27 2.88 Bangladesh 4.21 3.71 3.43 4.17 3.77 India 3.59 3.19 2.91 2.90 2.64 Table 11. Old age dependency ratio, regions of Asia and seven ANE countries
Old age dependency ratio 2025 Old age dependency ratio 2050 Var1 Var5 Var6 Var9 Var1 Var5 Var6
East Asia 0.197 0.210 0.215 0.226 0.307 0.383 0.450 Southeast Asia 0.113 0.124 0.130 0.138 0.193 0.262 0.323 South Asia 0.102 0.112 0.116 0.122 0.163 0.216 0.261
South Korea 0.211 0.226 0.232 0.246 0.326 0.417 0.488 Indonesia 0.112 0.123 0.130 0.139 0.185 0.260 0.328 Phiippines 0.096 0.105 0.108 0.117 0.160 0.216 0.255 Thailand 0.164 0.178 0.184 0.190 0.295 0.382 0.457 Egypt 0.102 0.112 0.117 0.127 0.159 0.221 0.270 Bangladesh 0.077 0.084 0.087 0.095 0.157 0.213 0.260 India 0.112 0.123 0.128 0.133 0.173 0.232 0.283
Old age dependency ratio is calculated as population 65 and older divided by population 15-64. Var1 is high fertility and mortality variant; Var5 is medium fertility and mortality variant, Var6 is low fertility and medium mortality variant; Var9 is low fertility and mortality variant. Table 12. Total dependency ratio, regions of Asia and seven ANE countries 1 2 3 4 5 6 7 8 9 Fertility variant: High Med Low High Med Low High Med Low Mortality variant: High High High Med Med Med Low Low Low Dependency ratio in 2025 East Asia 0.506 0.466 0.419 0.514 0.474 0.428 0.531 0.495 0.446 Southeast Asia 0.521 0.458 0.392 0.522 0.460 0.395 0.542 0.471 0.408 South Asia 0.539 0.470 0.398 0.540 0.472 0.401 0.559 0.488 0.424
South Korea 0.497 0.470 0.429 0.505 0.477 0.437 0.529 0.501 0.462 Indonesia 0.520 0.454 0.385 0.521 0.456 0.388 0.542 0.464 0.396 Phiippines 0.524 0.458 0.385 0.524 0.458 0.387 0.541 0.476 0.416 Thailand 0.496 0.450 0.398 0.497 0.453 0.402 0.518 0.468 0.422 Egypt 0.532 0.460 0.385 0.535 0.464 0.389 0.546 0.471 0.403 Bangladesh 0.501 0.429 0.357 0.498 0.428 0.357 0.533 0.462 0.400 India 0.527 0.455 0.381 0.530 0.459 0.385 0.540 0.466 0.398 Dependency ratio in 2050 East Asia 0.642 0.613 0.600 0.673 0.649 0.643 - - - Southeast Asia 0.595 0.546 0.511 0.613 0.570 0.544 - - - South Asia 0.570 0.508 0.457 0.579 0.522 0.477 - - -
South Korea 0.662 0.642 0.637 0.701 0.687 0.691 - - - Indonesia 0.595 0.545 0.510 0.614 0.573 0.548 - - - Phiippines 0.555 0.498 0.445 0.572 0.521 0.476 - - - Thailand 0.664 0.636 0.628 0.683 0.660 0.659 - - - Egypt 0.563 0.503 0.453 0.583 0.529 0.488 - - - Bangladesh 0.572 0.510 0.459 0.579 0.523 0.481 - - - India 0.576 0.515 0.467 0.587 0.531 0.490 - - -
The total dependency ratio is calculated as the population under 15 or 65 and older divided by the population 15-64. Table 13. Estimated number of adults and children with HIV/AIDS, end of 1997
Number of Percentage Number of adults (15-49) of adults children Japan 6,800 0.01 <100 South Korea 3,100 0.01 <100 Indonesia 51,000 0.05 960 Philippines 23,000 0.06 620 Thailand 770,000 2.23 14000 Bangladesh 21,000 0.03 27 India 4,100,000 0.82 48000 Egypt 8,100 0.03 27
Source: UN AIDS, various. Table 14. Migrant populations, proportion of foreign born in population, and the growth rate of migrant population, 1965, 1975, 1985, and 1990
1965 1975 1985 1990
World total 75,214 84,494 105,194 119,761
Asia 31,429 29,662 38,731 43,018 Foreign born East and Southeast 8,136 7,723 7,678 7,931 population (1000s) South-central 18,610 15,565 19,243 20,782
North America 12,695 15,042 20,460 23,895 Oceania 2,502 3,319 4,106 4,675 World total 2.3 2.1 2.2 2.3
Asia 1.7 1.3 1.4 1.4 Foreign born East and Southeast 0.7 0.5 0.5 0.4 percentage of South 2.8 1.9 1.8 1.8 total population (%) North America 6 6.3 7.8 8.6
Oceania 14.4 15.6 16.9 17.8 1965-1975 1975-1985 1985-1990 1965-1990 World total 1.2 2.2 2.6 1.9
Annual rate of Asia -0.6 2.7 2.1 1.3 Change in foreign born East population and Southeast -0.5 -0.1 0.6 -0.1 (%) South -1.8 2.1 1.5 0.4
North America 1.7 3.1 3.1 2.6
Oceania 2.8 2.1 2.6 2.5
Source: U.N., 1998. World Population Monitoring 1997 (New York: U.N.) Table 15. Net migration, net migration rate, population growth rate, and contribution of net migration to population growth
Number of net migrants (1000s) Net migration rate, per 1000 (A) 2000-05 2020-25 2045-50 2000-05 2020-25 2045-50
Asia -5,463 -3,941 -4,021 -0.3 -0.2 -0.2 East Asia -624 -820 -900 -0.1 -0.1 -0.1 Southeast Asia -1,556 -1,295 -1,390 -0.6 -0.4 -0.4 South Asia -3,740 -1,986 -1,891 -0.5 -0.2 -0.2
Japan 0 0 0 0 0 0 South Korea -100 0 0 -0.4 0 0 The Philippines -600 -500 -500 -1.5 -0.9 -0.8 Thailand -100 0 0 -0.3 0 0 Indonesia -900 -800 -890 -0.8 -0.6 -0.6 India -699 -686 -691 -0.1 -0.1 -0.1 Bangladesh -600 -500 -500 -0.9 -0.6 -0.5 Egypt -150 -150 -150 -0.4 -0.3 -0.3
North America 4,652 4,652 4,652 3 2.6 2.4 Oceania 419 437 436 2.7 2.3 1.9
Population growth rate, per 1,000 (B) Percentage of population growth due to migration (contribution) (A*100)/B 2000-05 2020-25 2045-50 2000-05 2020-25 2045-50
Asia 12.3 7.7 2.4 -2.5 -2.6 -8.3 East Asia 7.1 3 -2.7 -1.4 -3.4 3.8 Southeast Asia 13.3 9 3.5 -4.5 -4.4 -11.4 South Asia 15.8 10.2 4.8 -3.2 -2 -4.2
Japan 1.2 -4.5 -6.4 0 0 0 South Korea 7.2 2.5 -3.2 -5.6 0 0 The Philippines 18.8 11.1 5 -8 -8.1 -15.9 Thailand 8.3 4.9 -2 -3.6 0 0 Indonesia 12.2 8.3 3.4 -6.6 -7.2 -17.6 India 14.1 9 3.7 -0.7 -1.1 -2.7 Bangladesh 16.9 9.8 4.6 -5.3 -6.1 -10.9 Egypt 17 11 5.4 -2.4 -2.7 -5.6
North America 7.3 5.4 2 41.4 48.1 120 Oceania 12 8.8 4.7 22.4 26.2 40.3
Source: U.N. 1999. The percentage of population growth due to migration is calculated as the ratio of net migration rate divided by population growth rate. If both rates are positive, net migration contributes that percentage so as to increase population growth, whereas if net migration rates are negative, net migration contributes that percentage so as to decrease population growth. Table 16. Immigrants admitted to US per 1000 persons in the sending country, 1996-97
Age World Korea The Philippines India Bangladesh
Total 0.15 0.36 0.77 0.04 0.07 0-4 0.06 0.53 0.18 0.01 0.04 5-9 0.09 0.18 0.32 0.02 0.03 10-14 0.14 0.36 0.53 0.03 0.03 15-19 0.17 0.46 0.82 0.04 0.05 20-24 0.16 0.17 0.56 0.04 0.11 25-29 0.24 0.4 1.03 0.08 0.16 30-34 0.24 0.36 1.13 0.07 0.13 35-39 0.19 0.42 0.99 0.05 0.1 40-44 0.16 0.56 1.01 0.06 0.08 45-49 0.16 0.48 1.15 0.06 0.06 50-54 0.15 0.32 1.3 0.07 0.07 55-59 0.15 0.24 1.7 0.08 0.09 60-64 0.15 0.24 2.03 0.08 0.11 65-69 0.14 0.26 2.08 0.07 0.09 70-74 0.11 0.22 1.67 0.05 0.06 75-79 0.09 0.21 1.08 0.04 0.04 80+ 0.07 0.17 0.56 0.03 0.02
Source: U.S. Department of Justice, Immigration and Naturalization Service, 1997 and 1998. Statistical Yearbook of the Immigration and Naturalization Service. This is calculated as the ratio, of the average number of immigrants in each age group in 1996 and 1997 divided by the sending country’s population in that age group in 1995. Table 17. Macroeconomic indicators of selected Asian economies, 1994-98.
1994 1995 1996 1997 1998
GDP growth rate (%) Indonesia 7.5 8.2 8 4.5 -13.2 Korea 8.3 8.9 6.8 5 -5.8 Thailand 9 8.9 5.9 -1.8 -10 Philippines 4.4 4.7 5.8 5.2 -0.5 Inflation rate (%) Indonesia 8.6 9.4 8 6.7 57.7 Korea 6.2 4.5 4.9 4.5 7.5 Thailand 5 5.8 5.8 5.7 8.1 Philippines 8.4 8 9 6 9.7 Unemployment rate (%) Indonesia 4.4 7.2 4.9 4.7 5.5 Korea 2.4 2 2 2.6 6.8 Thailand 1.3 1.1 1.1 0.9 5.3 Philippines 8.4 8.4 7.4 7.9 9.6
Source: IMF, International Financial Statistics Database. ADB, Statistical Database System. Table 18. National health expenditures as percentage of GDP, 1997
Total Public Sector Private Sector
Bangladesh 4.9 2.3 2.6 Egypt 3.7 1.0 2.7 India 5.2 0.7 4.5 Indonesia 1.7 0.6 1.1 Japan 7.1 5.7 1.4 Philippines 3.4 1.6 1.8 South Korea 6.7 2.5 4.2 Thailand 5.7 1.9 3.8
Source : WHO Table 19. National health expenditures as percentage of GDP
Year Total Public Sector Private Sector
Bangladesh 1995 1995 2.4 1.2 1.3 Egypt 1995 1995 3.7 1.6 2.1 India 1991 1991 5.6 1.2 4.4 Indonesia 1994 1994 1.8 0.7 1.2 Japan 1995 1995 7.2 5.6 1.6 Philippines 1991 1991 2.4 1.3 1 South Korea 1992 1992 5.4 1.8 3.6 Thailand 1992 1992 5.3 1.4 3.9
Source : Bos, et al. 1999 Table 20. National health expenditure as percentage of GDP 1960-1996, G7 countries, Australia and South Korea
1960 1970 1980 1990 1997
Australia 4.9 5.7 7.3 8.2 8.4 Canada 5.4 7.0 7.2 9.2 9.2 France 4.2 5.8 7.6 8.9 9.6 Germany 4.8 6.3 8.8 8.7 10.7 Italy 3.6 5.2 7.0 8.1 7.6 Japan 3.0 4.6 6.5 6.1 7.2 North Korea 2.3 3.7 5.2 6.0 United Kingdom 3.9 4.5 5.6 6.0 6.8 United States 5.2 7.3 9.1 12.6 13.9
Source: OECD Health Data 1999 Table 21. Estimated income elasticities form health care expenditure data
Author Data Country/Region Year Estimated elasticity
Macro Studies Newhouse 1977 Cross-section 13 developped countries 1972 1.31 Parkin, et al. 1987 Cross-section 18 OECD countries 1980 0.9
Bos, et al. 1999 Cross-section 130 countries 1994 1.33 Fogel 1999 Long-run time-series United States 1875-1995 1.6 Hitiris and Posnett 1992 Panel 20 OECD countries 1960-1987 1 Murty and Ukpolo 1994 Timeseries USA 1960-1987 0.77 O'Connell 1996 Panel OECD contries 1975-1990 <1 Saez and Murillo 1994 Panel 10 OECD countries 1960-1990 1
Micro Studies Manning et al. 1986 Panel USA 1974-1977 0.0 Russo, et al. Cross-section Indonesia 1987 1.65 Russo and Herin 1993 Cross-section Philippines 1985 1.36 Russo 1992 Cross-section Thailand 1988 1.85 Table 22. Tuberculosis among populations 60 and older, 1990
Incidence Rate (per 100,000) Prevalence Rate (per 100,000)
Males Females Males Females
India 934.0 337.0 2,247.0 672.0 China 364.9 144.4 958.0 316.0 Other Asia and Islands 701.3 488.5 1,866.0 1,157.0 Middle Eastern cresent 236.0 153.0 395.0 236.0 Latin America and the Caribbean 237.0 121.0 374.0 192.0 Sub-Saharan Africa 393.0 295.0 923.0 699.0 Former socialist economies of Europe 84.7 29.4 43.9 14.9 Established market economies 57.5 21.5 14.4 5.3
Source: Murray and Lopez 1996, Global Health Statistics Table 23. Out-of-pocket payments as percentage of total health expenditure: South Korea
Out-of -pocket payments
1985 60.1 1986 61.8 1987 59.5 1988 56.5 1989 56.4 1990 53.0 1991 55.8 1992 60.3 1993 55.2 1994 53.3 1995 52.0
Source: OECD Health Data 1999 Table 24. Average retirement age among men, OECD countries 1950-1995
1950 1960 1970 1980 1990 1995 Decrease 1995-60
Australia 66.0 66.1 65.0 62.4 62.7 61.8 -4.3 Austria 66.4 63.9 62.7 60.1 58.7 58.6 -5.3 Belgium 64.8 63.3 62.6 61.1 58.3 57.6 -5.6 Canada 66.7 66.2 65.0 63.8 62.8 62.3 -3.9 Denmark 67.1 66.7 66.3 64.5 63.3 62.7 -4.0 Finland 66.8 65.1 62.7 60.1 59.6 59.0 -6.1 France 66.1 64.5 63.5 61.3 59.6 59.2 -5.3 Germany 65.7 65.2 65.3 62.2 60.3 60.5 -4.7 Greece 68.2 66.5 65.6 64.9 62.3 62.3 -4.2 Iceland 68.9 68.8 66.7 69.3 68.9 69.5 0.7 Ireland 68.3 68.1 67.5 66.2 64.0 63.4 -4.8 Italy 66.9 64.5 62.6 61.6 60.9 60.6 -3.8 Japan 66.7 67.2 67.7 67.2 66.5 66.5 -0.7 Luxembourg 65.8 63.7 62.5 59.0 57.6 58.4 -5.2 Netherlands 66.4 66.1 63.8 61.4 59.3 58.8 -7.3 New Zealand 65.8 65.1 64.7 62.9 62.2 62 -3.1 Norway 67.6 67.0 66.5 66.0 64.6 63.8 -3.2 Portugal 67.8 67.5 67.2 64.7 63.9 63.6 -4.0 Spain 68.1 67.9 65.2 63.4 61.6 61.4 -6.5 Sweden 66.8 66.0 65.3 64.6 63.9 63.3 -2.7 Switzerland 67.7 67.3 66.7 65.5 64.8 64.6 -2.7 Turkey 69.1 68.7 68.0 64.9 63.5 63.6 -5.2 United Kingdom 67.2 66.2 65.4 64.6 63.2 62.7 -3.5 United States 66.9 66.5 65.4 64.2 64.1 63.6 -2.9
Source: Blondal and Scarpetta (1998). Table 25. Median age at retirement, Asiane workers ages 40 and over, Asia 1950-2010
1950 1960 1970 1980 1990 1995 2000 2010
Asia na 66.7 65.7 64.2 63.6 63.2 62.8 62.1 East Asia 66.1 65.1 64.3 62.8 62.3 62.0 61.6 61.1 Southeast Asia na na na 66.7 65.5 64.9 64.4 63.5 South Asia na na na 67.0 66.2 65.5 65.0 64.1
South Korea na na 65.1 63.2 62.6 62.4 62.2 61.8 Indonesia na na na na 66.3 65.4 64.6 63.4 Philippines na na na na 66.6 66.1 65.6 64.9 Thailand na 65.1 64.5 63.8 63.2 63.0 62.8 62.5 Bangladesh na na na na 67.9 67.1 66.4 65.3 India na na na 67.6 66.6 66.1 65.5 64.4 Egypt na na 64.6 62.9 62.3 62.0 61.8 61.4
Source: ILO (1996). "na" indicates the median age at retirement exceeds age 67. Table 26. Main reasons for leaving the job, men aged 55-64 in 1995, EU
AustriaBelgiumGermanyFinlandFranceDenmarkItalyLuxembourgNetherlandPortugal Spain Sweden UK
Dismissed or made 5.1 3.7 23.4 24.1 10.7 9.5 2.0 0.0 7.9 1.0 10.2 30.2 22.0 redundant Job of limited duration 0.2 0.7 7.0 3.8 1.5 0.7 1.7 0.5 0.1 0.4 11.1 8.2 3.6 has ended Personal or family 0.2 1.3 0.2 0.0 0.4 0.6 1.2 0.2 1.9 0.0 0.2 2.0 1.6 responsibilites Own illness or disability 2.6 7.7 9.5 25.0 7.3 22.9 5.2 16.6 15.6 2.1 18.3 7.0 22.8 Early retirement 49.0 30.6 37.2 0.0 16.9 33.1 9.2 29.1 42.9 2.3 13.0 25.9 14.7 Normal retirement 30.2 19.6 2.3 11.7 38.6 10.9 53.4 31.7 0.0 1.2 17.8 12.5 4.8 Other reasons 12.8 36.4 20.3 35.5 24.5 22.3 27.5 22.0 31.6 93.1 29.5 14.2 30.6
Total 100 100 100 100 100 100 100 100 100 100 100 100 100
Note : 1. Refers to personsaged 55-64 who are not in the labour force, but who had been in the labour force in the 8 years preceeding the survey. 2. Other reasons include: education and training, compulsory military or community service and other reasons. Source: The European Union Labour Survey 1995, quoted from Auer and Fortuny (2000). Table 27. Statutory Retirement Age
Statutory Retirement Age Statutory Retirement Age Men Women Men Women
China 60 55 United Kingdom 65 60 Japan 65 65 Italy 65 60 Republic of Korea 60 60 Spain 65 65 Bangladesh .. .. Austria 65 60 India 55 55 Belgium 65 61 Pakistan 60 55 France 60 60 Sri Lanka 55 50 Germany 63 63 Indonesia 55 55 Liechtenstein 65 62 Philippines 60 60 Luxembourg 65 65 Singapore 55 55 Monaco 65 65 Thailand .. .. Netherlands 65 65 Vietnam 60 55 Switzerland 65 62 Denmark 67 67 Mexico 65 65 Finland 65 65 Chile 65 60 Iceland 67 67 Canada 65 65 Norway 67 67 USA 65 65 Sweden 65 65 Australia 65 61
Source: SSA (1999) and UN (1999). Table 28. Living arrangements of the elderly Family households Non-family households Household head/spouse Parent of Other Multi- One Combined Intact Not intact head member Combined person person Institution Males Japan, 1970 0.936 0.575 0.084 0.259 0.018 0.063 0.005 0.030 0.028 Japan, 1975 0.928 0.580 0.074 0.257 0.017 0.072 0.002 0.038 0.032 Japan, 1980 0.926 0.606 0.081 0.229 0.010 0.074 0.002 0.039 0.034 Japan, 1985 0.922 0.628 0.087 0.197 0.009 0.078 0.001 0.042 0.035 Japan, 1990 0.917 0.655 0.091 0.163 0.008 0.083 0.001 0.048 0.034 Japan, 1995 0.904 0.719 0.046 0.132 0.007 0.096 0.003 0.061 0.032 Korea, 1970 0.998 0.557 0.085 0.315 0.041 0.003 0.000 0.002 0.000 Korea, 1980 0.978 0.633 0.077 0.260 0.008 0.021 0.003 0.015 0.003 Korea, 1985 0.974 0.601 0.125 0.234 0.014 0.026 0.001 0.021 0.004 Korea, 1990 0.936 0.691 0.060 0.175 0.010 0.064 0.004 0.033 0.027 Thailand, 1970 0.976 0.651 0.137 0.085 0.103 0.024 0.001 0.023 na Thailand, 1980 0.973 0.660 0.143 0.145 0.025 0.027 0.003 0.024 na Thailand, 1990 0.923 0.654 0.140 0.102 0.027 0.077 0.002 0.033 0.042 Indonesia, 1980 0.974 0.646 0.158 0.138 0.033 0.026 0.001 0.025 na Indonesia, 1990 0.970 0.653 0.150 0.117 0.050 0.030 0.001 0.029 na Females Japan, 1970 0.900 0.205 0.060 0.580 0.055 0.100 0.004 0.069 0.027 Japan, 1975 0.871 0.221 0.051 0.545 0.055 0.129 0.002 0.090 0.037 Japan, 1980 0.844 0.247 0.041 0.523 0.032 0.156 0.002 0.110 0.045 Japan, 1985 0.818 0.272 0.037 0.479 0.030 0.182 0.002 0.129 0.051 Japan, 1990 0.797 0.320 0.033 0.416 0.028 0.203 0.001 0.148 0.054 Japan, 1995 0.783 0.352 0.064 0.345 0.022 0.217 0.002 0.162 0.053 Korea, 1970 0.997 0.138 0.058 0.694 0.107 0.017 0.000 0.000 0.017 Korea, 1980 0.923 0.145 0.084 0.678 0.016 0.077 0.006 0.069 0.002 Korea, 1985 0.898 0.160 0.038 0.641 0.060 0.102 0.003 0.099 0.000 Korea, 1990 0.857 0.185 0.097 0.538 0.037 0.142 0.007 0.131 0.004 Thailand, 1970 0.961 0.413 0.219 0.140 0.190 0.039 0.002 0.037 na Thailand, 1980 0.955 0.407 0.254 0.253 0.041 0.045 0.003 0.042 na Thailand, 1990 0.935 0.301 0.309 0.269 0.056 0.061 0.003 0.055 0.003 Indonesia, 1980 0.852 0.190 0.052 0.496 0.115 0.148 0.004 0.144 na Indonesia, 1990 0.845 0.232 0.050 0.438 0.125 0.155 0.002 0.153 na Notes: Based on unpublished census tabulations. For Singapore and Indonesia proportions are of the non-institutionalized population. Source: Natividad and Cruz (1997), Knodel and Debavalya (1997), Knodel and,Chayovan (1997), Andrews and Hennink (1992), Casterline et. al. (1991), Domingo and Casterline (1992), Kim (1997), and Chan (1997). Table 28. Non-Family Living Arrangements for the Elderly (65 and older) in Asia
Non-family households Non-family households Family Multi- One Family Multi- One Households Combined person person Institutions Households Combined person person Institutions Males Females Japan, 1970 0.936 0.063 0.005 0.030 0.028 0.900 0.100 0.004 0.069 0.027 Japan, 1975 0.928 0.072 0.002 0.038 0.032 0.871 0.129 0.002 0.090 0.037 Japan, 1980 0.926 0.074 0.002 0.039 0.034 0.844 0.156 0.002 0.110 0.045 Japan, 1985 0.922 0.078 0.001 0.042 0.035 0.818 0.182 0.002 0.129 0.051 Japan, 1990 0.917 0.083 0.001 0.048 0.034 0.797 0.203 0.001 0.148 0.054 Japan, 1995 0.904 0.096 0.003 0.061 0.032 0.783 0.217 0.002 0.162 0.053
Korea, 1970 0.998 0.003 0.000 0.002 0.000 0.997 0.017 0.000 0.000 0.017 Korea, 1980 0.978 0.021 0.003 0.015 0.003 0.923 0.077 0.006 0.069 0.002 Korea, 1985 0.974 0.026 0.001 0.021 0.004 0.898 0.102 0.003 0.099 0.000 Korea, 1990 0.936 0.064 0.004 0.033 0.027 0.857 0.142 0.007 0.131 0.004
Thailand, 1970 0.976 0.024 0.001 0.023 na 0.961 0.039 0.002 0.037 na Thailand, 1980 0.973 0.027 0.003 0.024 na 0.955 0.045 0.003 0.042 na Thailand, 1990 0.923 0.077 0.002 0.033 0.042 0.935 0.061 0.003 0.055 0.003
Indonesia, 1980 0.974 0.026 0.001 0.025 na 0.852 0.148 0.004 0.144 na Indonesia, 1990 0.970 0.030 0.001 0.029 na 0.845 0.155 0.002 0.153 na
Philippines, 1970 0.952 0.048 0.002 0.046 na 0.912 0.088 0.004 0.084 na Philippines, 1980 0.965 0.035 0.002 0.034 na 0.948 0.052 0.002 0.050 na Philippines, 1990 0.962 0.038 0.002 0.036 na 0.926 0.074 0.005 0.069 na
Bangladesh, 1981 na na na na 0.041 na na na na 0.022 Bangladesh, 1991 na na na na 0.015 na na na na 0.005
Notes: Based on unpublished census tabulations. For Singapore and Indonesia proportions are for the non-institutionalized population. Data for Bangladesh are drawn from published census reports (Bangladesh, 1984, 1994).
Table 29. Non-Family Living Arrangements for the Elderly (65 and older) in Asia
Non-family households Non-family households Family Multi- One Family Multi- One Households Combined person person Institutions Households Combined person person Institutions Males Females Japan, 1970 0.936 0.063 0.005 0.030 0.028 0.900 0.100 0.004 0.069 0.027 Japan, 1975 0.928 0.072 0.002 0.038 0.032 0.871 0.129 0.002 0.090 0.037 Japan, 1980 0.926 0.074 0.002 0.039 0.034 0.844 0.156 0.002 0.110 0.045 Japan, 1985 0.922 0.078 0.001 0.042 0.035 0.818 0.182 0.002 0.129 0.051 Japan, 1990 0.917 0.083 0.001 0.048 0.034 0.797 0.203 0.001 0.148 0.054 Japan, 1995 0.904 0.096 0.003 0.061 0.032 0.783 0.217 0.002 0.162 0.053
Korea, 1970 0.998 0.003 0.000 0.002 0.000 0.997 0.017 0.000 0.000 0.017 Korea, 1980 0.978 0.021 0.003 0.015 0.003 0.923 0.077 0.006 0.069 0.002 Korea, 1985 0.974 0.026 0.001 0.021 0.004 0.898 0.102 0.003 0.099 0.000 Korea, 1990 0.936 0.064 0.004 0.033 0.027 0.857 0.142 0.007 0.131 0.004
Thailand, 1970 0.976 0.024 0.001 0.023 na 0.961 0.039 0.002 0.037 na Thailand, 1980 0.973 0.027 0.003 0.024 na 0.955 0.045 0.003 0.042 na Thailand, 1990 0.923 0.077 0.002 0.033 0.042 0.935 0.061 0.003 0.055 0.003
Indonesia, 1980 0.974 0.026 0.001 0.025 na 0.852 0.148 0.004 0.144 na Indonesia, 1990 0.970 0.030 0.001 0.029 na 0.845 0.155 0.002 0.153 na
Philippines, 1970 0.952 0.048 0.002 0.046 na 0.912 0.088 0.004 0.084 na Philippines, 1980 0.965 0.035 0.002 0.034 na 0.948 0.052 0.002 0.050 na Philippines, 1990 0.962 0.038 0.002 0.036 na 0.926 0.074 0.005 0.069 na
Bangladesh, 1981 na na na na 0.041 na na na na 0.022 Bangladesh, 1991 na na na na 0.015 na na na na 0.005
Notes: Based on unpublished census tabulations. For Singapore and Indonesia proportions are for the non-institutionalized population. Data for Bangladesh are drawn from published census reports (Bangladesh, 1984, 1994). Table 30. Percentage living with children by parents' age, ASEAN survey
Philippines Singapore Taiwan Thailand (1984) (1986) (1989) (1986) Total 60-64 84.6 94.2 77.5 84.0 65-69 82.1 93.2 76.2 80.4 70-74 81.0 93.3 72.1 76.5 75+ 76.8 90.6 73.1 74.9 Male 60-69 86.2 94.3 76.3 83.2 70+ 80.8 89.7 68.5 72.7 Female 60-69 80.9 93.4 77.9 81.8 70+ 77.1 93.7 76.7 77.4
Source : Casterline et al. (1991). Table 31. Proportion of older adults living alone by age and sex, selected Asian countries
Males Females Country (year) 55-59 60-64 65-69 70-74 75-79 80-84 85+ 55-59 60-64 65-69 70-74 75-79 80-84 85+
Japan (1970) 0.020 0.024 0.026 0.036 0.033 0.041 0.023 0.062 0.068 0.077 0.074 0.069 0.057 0.036 Japan (1975) 0.027 0.028 0.036 0.040 0.042 0.046 0.040 0.080 0.093 0.098 0.100 0.092 0.076 0.050 Japan (1980) 0.027 0.031 0.035 0.041 0.045 0.056 0.051 0.083 0.110 0.123 0.123 0.105 0.091 0.061 Japan (1985) 0.037 0.034 0.037 0.046 0.050 0.058 0.056 0.078 0.109 0.140 0.145 0.139 0.102 0.069 Japan (1990) 0.050 0.041 0.047 0.047 0.055 0.069 0.065 0.077 0.110 0.140 0.172 0.173 0.142 0.091
China (1990) 0.042 0.051 0.066 0.092 0.102 0.128 0.130 0.024 0.047 0.079 0.115 0.142 0.151 0.149 Taiwan (1990) 0.056 0.060 0.068 0.084 0.101 0.129 0.157 0.038 0.051 0.064 0.074 0.078 0.075 0.071
Korea (1970) 0.002 0.001 0.003 0.002 0.002 0.000 0.011 0.001 0.001 0.000 0.001 0.001 0.002 0.000 Korea (1980) 0.010 0.014 0.013 0.015 0.023 0.018 0.022 0.048 0.069 0.077 0.071 0.064 0.045 0.050 Korea (1985) 0.022 0.024 0.027 0.030 0.024 0.031 0.016 0.060 0.091 0.111 0.100 0.085 0.067 0.048 Korea (1990) 0.022 0.027 0.032 0.034 0.039 0.032 0.047 0.067 0.121 0.146 0.144 0.125 0.098 0.060
Philippines (1970)0.024 0.030 0.043 0.046 0.043 0.044 0.062 0.034 0.050 0.078 0.086 0.085 0.104 0.081 Philippines (1975)0.020 0.024 0.031 0.037 0.032 0.041 0.031 0.014 0.022 0.037 0.056 0.056 0.049 0.059 Philippines (1980)0.009 0.013 0.025 0.036 0.049 0.051 0.078 0.012 0.022 0.034 0.055 0.073 0.106 0.073 Philippines (1990)0.019 0.023 0.025 0.039 0.043 0.061 0.046 0.021 0.036 0.059 0.062 0.086 0.088 0.080
Thailand (1970) 0.017 0.019 0.029 0.037 0.041 0.054 0.038 0.038 0.057 0.074 0.088 0.114 0.110 0.115 Thailand (1980) 0.017 0.025 0.028 0.026 0.031 0.031 0.045 0.024 0.031 0.053 0.063 0.061 0.070 0.055 Thailand (1990) 0.022 0.024 0.032 0.031 0.037 0.038 0.053 0.027 0.042 0.049 0.055 0.069 0.055 0.056
Indonesia (1976) 0.012 0.013 0.032 0.029 0.028 0.043 0.012 0.080 0.108 0.130 0.148 0.192 0.183 0.130 Indonesia (1980) 0.011 0.017 0.020 0.029 0.033 0.037 0.040 0.064 0.107 0.120 0.139 0.133 0.132 0.107 Indonesia (1990) 0.014 0.020 0.023 0.033 0.041 0.048 0.047 0.058 0.105 0.125 0.151 0.133 0.160 0.128
Source: Values tabulated from primary census data for each country. Table 32. Household size, children and adults per household, Asian countries.
Average Children Adults Average Children Adults Year of Household per per Year of Household per per Census Size HouseholdHousehold Census Size HouseholdHousehold
Bangladesh Malaysia 1974 5.64 2.71 2.93 1970 5.50 2.49 3.01 1981 5.78 2.69 3.09 1980 5.20 2.08 3.12 1991 5.48 2.47 3.01 Myanmar India 1983 5.49 2.12 3.37 1971 5.60 2.34 3.26 Nepal 1981 5.60 2.21 3.39 1971 5.53 2.38 3.15 1991 5.57 2.03 3.54 1981 5.80 2.40 3.40 Indonesia 1991 5.60 2.36 3.24 1971 4.87 2.14 2.73 Pakistan 1980 4.85 1.98 2.87 1981 6.70 2.91 3.79 1985 4.57 1.80 2.77 Philippines 1990 4.51 1.65 2.86 1970 5.95 2.72 3.23 1995 4.27 1.44 2.82 1975 5.94 2.61 3.33 Japan 1980 5.59 2.35 3.24 1950 4.97 1.79 3.18 1990 5.32 1.94 3.38 1955 4.97 1.72 3.25 Singapore 1960 4.54 1.43 3.11 1970 5.60 2.29 3.31 1965 4.05 1.10 2.95 1980 4.90 1.33 3.57 1970 3.69 0.93 2.76 1990 4.20 0.95 3.25 1975 3.45 0.87 2.58 Taiwan 1980 3.33 0.81 2.52 1956 5.70 2.51 3.19 1985 3.22 0.71 2.51 1966 5.68 2.26 3.42 1990 2.99 0.57 2.42 1970 5.52 2.27 3.25 1995 2.97 0.47 2.50 1975 5.25 1.94 3.31 South Korea 1980 4.80 1.55 3.25 1955 5.50 2.34 3.16 1990 3.99 1.11 2.88 1960 5.70 2.32 3.38 Thailand 1966 5.49 2.51 2.98 1960 5.68 2.47 3.21 1970 5.40 2.37 3.03 1970 5.82 2.62 3.20 1975 5.10 1.99 3.11 1980 5.20 2.04 3.16 1980 4.60 1.59 3.01 1990 4.40 1.29 3.11 1985 4.22 1.26 2.96 Viet Nam 1990 3.82 0.98 2.84 1979 5.00 2.25 2.75
Source: Population censuses, details available from authors. Table 33. Types of schemes providing cash benefits to the aged, disabled and/or survivors, 1999
No Contributory Non-Contributory Mandatory Mandatory savings scheme Flat-rateEarnings- Means- Flat-rate private public private related related universal pensions Africa 2 2 35 2 1 1 6 0 Egypt - - yes - - - -
Asia 2 5 26 9 1 1 6 0 Japan - yes yes - - - - - South Korea - - yes - - - - - Philippines - - yes - - - - - Thailand - - yes - - - - - Indonesia ------yes - Bangladesh yes ------India ------yes -
Latin/Carribean 1 1 34 4 0 0 0 8
Europe 0 15 36 19 2 4 2 2
North America 0 0 2 2 1 0 0 0
Oceania 0 0 3 1 1 1 6 0
World Total 5 23 136 37 6 7 20 10
Source: Social Security Administration (1999)
< Definition of Terms > Contributory flat-rate pension is a pension of uniform among or based on years of service or residence but independent of earnings that is financed by payroll-tax contributions from employees and/or employers. Contributory earnings-related pension is a pension based on earnings which is financed by payroll-tax contributions from employees and/or employers.
Non-contributory means-tested pension is a pension paid to eligible persons whose own of family income or assets fall below designated levels that is generally financed through government contributions with no contribution from employer or employees.
Non-contributory flat-rate universal pension is a pension of uniform amount of based years of service but independent of earnings that is paid to residents who meet age or disability that is financed with no contributions from employers or employees.
Mandatory private pension system is a system requiring employers by law, to provide private/occupational pensions.
Mandatory saving system is a compulsory defined-contribution pension system which pays benefits either as a lump-sum or as an annuity based on employee and in some cases employer contributions and returns on investment funds including publicly managed provident funds and privately managed systems. Table 34. Coverage of schemes providing cash benefits to the old aged, disabled and/or survivals Countries (coverage rate Coverage of schemes (1999) Special schemes in 1992)
Japan (n.a.) National pension program: All residents Public employees, private school teacher, aged 20-59. Voluntary coverage for agriculture, fishery and forest sector residents aged 60-64 (employees’ pension insurance: employees of firm in industry and commerce)
South Korea (25.9%) National pension program: employers Public employees, private school teacher, and employees in workplace with 5 or military personnel more employees. The self-employed, farmers, and fishermen aged 18-59. Voluntary coverage for residents aged 60- 64
Philippines (52.6%) All private employees under 60. House Government employees, sugar industry helpers or self-employed earning at least workers, military personnel 1,000 pesos a month Thailand (n.a.) Employees of firm with 10 or more Civil servants, private school teacher workers. Voluntary coverage for the self- employed Indonesia (6.9%) Employees earning less than 5,000 Public employees, military personnel rupees a month from establishments with 10 or more workers or a payroll of Rp. 1 million or more a month
India (0.9%) Employees of establishments with 20 or Public employees, railway industry and more employees in 177 categories of coal-miners industries
Bangladesh (0.0%) None Public employees Malaysia (95.6%) Mandatory coverage for private sector Armed forces employees, non-pensionable public sector employees and foreign sector employees. Voluntary coverage for domestic workers, self-employed, and pensionable public sector employees
China (21.1%) Employees in state-run enterprises. Government and party employees, Private, collective and foreign companies cultural, scientific, and educational depend on local government regulations institutions
Egypt (n.a.) Employees aged 18 and over. Civil servants, casual, agricultural, and Government employees aged 16 and self-employed over
Source: Social Security Administration (1999), ILO (1995). Table 35. Percentages of social security expenditure and tax revenue, selected years
1972 1975 1980 1985 1990 1991 1992 1993 1994 1995 1996
(A) Public expenditure on social security (% of public expenditure)
Philippines 3 2 1 2 1 2 2 2 na na na South Korea 5 5 6 5 7 8 8 8 na na na Singapore 0 2 1 1 2 2 2 2 na na na
Canada na 30 26 27 29 32 na na na na na Denmark 32 36 36 33 38 39 na na na na na Luxembourg 46 46 47 49 45 48 47 na na na na Netherlands na 34 35 34 34 34 35 na na na na Spain 47 50 54 39 36 37 na na na na na Sweden 32 36 38 41 48 46 43 na na na na United States 27 31 28 24 20 21 22 22 na na na Argentina na na 19 19 22 22 na na na na na Brazil na 50 32 21 23 33 28 na na na na Chile 30 24 31 37 34 32 31 31 na na na
(B) Employer's and employee's social security contributions (% of tax revenue)
South Korea 0.7 0.8 1.1 1.5 4.5 4.9 5.5 8.2 7.5 7.5 8.8 Malaysia 0.1 0.5 0.4 0.5 0.8 0.8 0.9 1.1 1.1 1.2 1.1 Thailand 0.0 0.1 0.2 0.2 0.1 0.7 1.0 1.1 1.4 1.3 1.4 Egypt na 13.2 8.9 13.4 14.6 10.9 8.9 9.0 9.2 9.7 na
Canada na 10.0 10.4 14.5 14.7 16.7 18.4 19.8 18.9 na na Denmark 5.1 1.8 2.2 4.6 3.8 3.7 3.9 3.9 4.0 3.9 na Luxembourg 27.7 29.1 26.0 23.9 24.6 24.7 26.6 26.7 25.9 26.8 25.8 Netherlands na 37.2 36.3 39.6 35.5 36.1 37.2 36.8 40.6 41.6 39.7 Spain 38.8 44.6 48.0 41.0 38.1 37.8 37.8 38.9 39.5 38.6 na Sweden 21.6 26.8 33.2 27.7 31.1 31.2 37.3 34.3 33.1 36.5 37.1 United States 23.6 28.0 28.2 32.9 34.6 35.1 35.5 34.2 34.3 33.3 33.0 Argentina na na 16.7 27.1 43.3 44.1 47.0 37.3 36.8 34.8 29.9 Brazil na na 25.0 20.6 22.4 24.6 24.9 26.4 26.5 na na Chile 28.6 9.9 16.9 7.3 8.3 7.0 7.0 6.6 6.5 6.1 6.1
Source: World Bank (1999). World Bank Database.
Table 36. Percentage of old-age benefit and other social security expenditure in GDP, circa 1993
Total social security Pension Other social security
Bangladesh 0.02 0.00 0.02 Thailand 0.12 0.00 0.12 Philippines 3.01 0.78 2.23 Indonesia 0.06 0.01 0.05 India 0.32 0.22 0.10 Singapore 1.78 1.31 0.47 Malaysia 0.15 0.00 0.15 South Korea 2.18 0.14 2.04 Japan 17.88 8.06 9.82 China 2.55 1.63 0.92 Egypt 1.20 na na
USA 15.45 6.26 9.19 Switzerland 20.53 6.18 14.35 UK 21.60 4.49 17.11 Canada 22.84 4.59 18.25 Austria 25.60 11.00 14.60 Germany 26.33 10.10 16.23 Netherland 31.70 7.14 24.56 Sweden 40.05 17.72 22.33 Italy 12.40 11.41 0.99 Denamrk 32.10 8.69 23.41 Spain 22.60 8.80 13.80 Chile 22.67 17.64 5.03
Source: ILO, Social Security Department database.